key: cord-0005636-yg7salqt
authors: nan
title: Oral Sessions and Working Party
date: 2008-03-26
journal: Bone Marrow Transplant
DOI: 10.1038/bmt.2008.30
sha: c1ac8c7ed9023bcd28bb01e1e51a7c431d20e8d8
doc_id: 5636
cord_uid: yg7salqt

nan

Basic Science Award: O100 MT1-MMP and RECK inversely regulate haematopoietic progenitor cell egress A. Avigdor* (1) , Y. Vagima (2) , P. Goichberg (2) , O. Kollet (2) , S. Shivtiel (2) , M. Tesio (2) , A. Kalinkovich (2) , I. Petit (2) , O. Perl (1) , E. Rosenthal (1) , I. Resnick (3), I. Hardan (1) , A. Nagler (1) , T. Lapidot (2) (1)The Chaim Sheba Medical Center (Tel-Hashomer, IL) ; (2) The Weizmann institute of Science (Rehovot, IL);

(3)Hadassah Medical Center (Jerusalem, IL) Hematopoietic progenitor cell release to the circulation is the outcome of signals provided by cytokines, chemokines, adhesion molecules, and proteases. Yet, the mechanisms of progenitor cell egress during G-CSF mobilization are not fully understood. Membrane type-1 metalloproteinase (MT1-MMP) and its endogenous inhibitor, RECK, are established key regulators of tumor cell motility. We detected higher MT1-MMP and lower RECK expression on circulating human CD34+ progenitors and maturing leukocytes as compared to immature bone-marrow (BM) cells. MT1-MMP expression was more prominent on CD34+ cells obtained from PB of G-CSFtreated healthy donors whereas RECK was barely detected. G-CSF mobilization in NOD/SCID mice, previously engrafted with human cells, increased MT1-MMP and decreased RECK expression on human progenitors and maturing leukocytes, in a PI3K/Akt1-dependent manner, resulting in elevated MT1-MMP activity. Blocking MT1-MMP function impaired G-CSF mobilization, while RECK neutralization promoted egress of human CD34+ progenitors. Targeting MT1-MMP expression by SiRNA or blocking its function reduced the in-vitro SDF-1 induced migration of human progenitors via matrigel and impaired the BM homing capacity of transplanted human progenitors in NOD/SCID mice. In accordance, neutralization of RECK function facilitated the migration of human BM CD34+ cells in vitro. Furthermore, following G-CSF mobilization, we also observed a reduction of CD44 on human CD34+ progenitors in the BM of chimeric mice. This was accompanied by accumulation of CD44 cleaved products of molecular weights, expected for MT1-MMP activity, in the BM supernatants. Blocking MT1-MMP function in chimeric mice resulted in less cleavage of CD44 upon G-CSF mobilization, whereas in the absence of a mobilizing signal, increasing MT1-MMP activity by RECK Ab injection facilitated CD44 proteolysis on the BM cells. Finally, MT1-MMP expression correlated with the number of CD34+ cells, collected on the first apheresis day in consecutive healthy donors and patients mobilized with G-CSF. In conclusion, our results indicate that G-CSF inversely regulates MT1-MMP and RECK expression on CD34+ progenitors, resulting in net increase in MT1-MMP activity. MT1-MMP proteolysis of CD44 diminishes progenitor adhesion to BM components, leading to cell egress. These previously undefined cell autonomous changes in the course of G-CSF treatment might serve as target for new approaches to improve mobilization.

Morbidity and mortality associated with treatment-related organ toxicity is a major factor limiting success of allogeneic hematopoietic stem cell transplantation (HSCT). Conservative therapeutic strategies have been ineffective in part, resulting in high rates of progression or complete organ failure and death. Over the last decades, solid organ transplantation (SOT) has been increasingly used for the treatment of terminal organ failure in HSCT recipients. To date, information regarding the use of SOT as treatment attempts in patients after HSCT is limited. As well the risk factors accounting for the necessity of SOT after HSCT as well as the incidence and outcome of this therapy are not well defined. A questionnaire survey was carried out within EBMT centres. 107 centres participated in this survey, covering allogeneic HSCT between 1984 and 2005 . 31 cases of SOT were identified. In more detail, 13 liver-, 8 kidney-, 9 lung-, and one heart transplantations were performed in 25 different centres. Indications for liver transplantation were infections leading to cirrhosis (n=4), sinusoidal obstruction syndrome (n=3), and GvHD of the liver (n=2). Rejection of the transplanted liver or terminal organ failure occurred in one patient respectively. Other complications after liver transplantation were infections (n=2), bleedings (n=2) and kidney failure (n=2). Most kidney transplantations were performed because of chronic kidney failure due to drug toxicity (n=6). Transplant rejection and/ or kidney failure did not occur. Interestingly, two of the kidney donors were also stem cell donors for the transplant recipient. Lung transplantation was performed in all cases because of bronchiolitis obliterans and/ or GvHD which led to respiratory failure. Rejection occurred in 4 patients and in one patient terminal transplant failure occurred. Other major complications in the lung transplant recipients were kidney failure (n=3) and infections (n=2). Heart Transplantation was performed in one patient because of pre-terminal heart insufficiency due to drug treatment. In summary, very few SOT for terminal organ failures were performed. The overall survival of patients receiving an organ graft after HSCT was 72.5% at 5 years with a median follow up time of 23 months. Complications after SOT, like infections and organ rejection were frequent, but manageable. We conclude that SOT offers a viable therapeutic option for patients who develop terminal organ failure after HSCT.

Influence of immunisation timing on the response to conjugate-pneumococcal vaccine after allogeneic stem cell transplant: final results of the EBMT IDWP01 Trial C. Cordonnier* (1) , M. Labopin (2) , V. Chesnel (2) , P. Ribaud Background: Pneumococcal infections are causes of death after SCT. The efficacy of the polysaccharide 23-valent vaccine (PPV23) is limited before 6 mo after SCT, or if graftversus-host disease (GVHD) . Previous studies with the Wyeth heptavalent conjugate Prevnar® vaccine (PCV7), using different schedules after allogeneic SCT, showed a response around 65-85 % of the patients. However, the optimal timing of vaccination is yet not defined, and pneumococcal infection may occur early in the first months after SCT. Our objective was to show that the response to early (E) (3 mo) immunization is not inferior to a late (L) (9 mo) immunization. Methods: Patients ≥ 5 year old and at 3 months after allogeneic myeloablative SCT were randomized to receive 3 doses of PCV7 at 1 month interval, followed by a PPV23 6 months later, from 3 (E) or from 9 (L) months after transplant. The primary endpoint was the % of responders (≥ .15 µg/ml of each of the 7 PCV7 serotypes) 1 month after the 3rd dose of PCV7 (S3). Ab levels were blindly measured by ELISA. All patients were followed until 24 months after transplant, or until death, whichever occurred first. Results: 158 patients were randomized: 75 in the early (E), and 83 in the late (L) group. Most patients were adults with acute leukemia, transplanted from an HLA-identical donor. 114 patients were evaluable for the primary endpoint (E: 57, L: 57). The response rate was respectively 79% vs 82% at S3, and not inferior in the early, when compared to the late group (90% CI; 8.6) . However, at 24 months, significantly less E patients were still protected when compared to L patients (26/44; 59% s 35/42; 83%, p=.013). In the lack of difference in the response between groups, the 2 groups were pooled to analyse the impact of transplant characteristics on the percentage of responders at S3. Donor age (> 36y) and chronic GVHD were the only factors impairing the response in the multivariate analysis. Conclusion: The Ab response 1 month after 3 doses of PCV7 after allogeneic SCT is about 80% and non inferior when started at 3 than at 9 months. We therefore recommend starting immunization at 3 months to offer an earlier protection. However, the E vaccination offers a significantly lower protection at 2 years, suggesting the need for a boost during the second year after E immunization. The authors are grateful to the Safety Committee: D. Engelhard, P. Reusser, and P. Reinert

Depletion of the autoreactive immunological memory followed by autologous haemopoietic stem cell transplantation in patients with refractory SLE induces long-term remissions through de novo generation of a juvenile and tolerant immune system T. Alexander (1) , A. Thiel (2) , G. Massenkeil (1) , A. Sattler (1) , S. Kohler (2) , H. Mei (2) , H. Radtke (1) , G.R. Burmester (1) , A. Radbruch (2) , R. Arnold (1) , F. Hiepe* (1) (1)Charité -Universitätsmedizin Berlin (Berlin, DE) ; (2) German Arthritis Research Center (Berlin, DE) Clinical trials have indicated that immunoablation followed by autologous haemopoietic stem cell transplantation (ASCT) has the potential to induce clinical remission in patients with refractory systemic lupus erythematosus (SLE). To elucidate the mechanisms mediating the beneficial long-term clinical responses, we investigated the immune reconstitution in SLE patients receiving ASCT as part of a monocentric phase I/II clinical trial. Seven patients with SLE were evaluated during a long-term follow-up (median follow-up period 60 months) who were immunoablated with cyclophosphamide and rabbit antithymocyte globulin, followed by transplantation of purified autologous CD34+ haemopoietic stem cells. Previous failure of conventional immunosuppression, including cyclophosphamide, had been an inclusion criterion. Humoral immunity was evaluated, peripheral T and B lymphocytes were immunophenotyped and frequencies of T lymphocytes specific for distinct antigens of interest were assessed after short-term stimulation ex vivo. In all patients clinical and serological remission was observed, accompanied by disappearance of anti-dsDNA autoantibodies and protective antibodies from serum. One patient developed a relapse 18 months after ASCT. In the responding patients, CD31+ CD45RA+ CD4+ T cells, i.e. recent thymic emigrants, recurred with a doubling in absolute counts compared to agematched healthy controls until 4-yr post-transplant (p=0.014). Absolute numbers of CD4+ FoxP3+ regulatory T cells (Tregs) normalised after ASCT and TCR repertoires CD4+ T cells displayed a broad clonal diversity as compared to the pretransplant status. Early after ASCT, often high frequencies of virus-specific effector T cells were detected. Autoreactive T cells specific for nucleosomes or SmD1 were not detectable. A normal B cell compartment developed within 12 months after therapy, as compared to the pre-existing B cell deficiencies, which had included naive (IgD+) B cell lymphopenia (p=0.031), relative predominance of memory (IgD-) B cells (p=0.016) and expansion of CD27high CD20plasma blasts. Our data demonstrate that the long-term therapy-free clinical remissions observed in SLE patients after complete immunoablation and ASCT are accompanied by a loss of immunological memory and a fundamental reset of the immune system. Depletion of the autoreactive memory and reactivation of thymic education probably are the basis for regeneration of self-tolerance and clinical remission.

M. Themeli* (1), L. Petrikkos (2) , M. Waterhouse (2) , H. Bertz (2) , N. Zoumbos (1) , J. Finke (2) , A. Spyridonidis (1) (1)University of Patras Medical School (Rion-Patras, GR); (2)Freiburg University Medical Center (Freiburg, DE) We previously demonstrated frequent genomic alterations measured by microsatellite instability (MSI) in non-neoplastic epithelial tissues of pts who underwent allogeneic hematopoietic cell transplantation (HCT) but not in pts after autologous HCT (Blood 2006; 107:3389) . Confirmation in larger independent patient cohort and in an in vitro system was needed. 176 buccal samples from 71 unselected pts obtained 30-3722 days (median 322) after allogeneic HCT were analysed for MSI. Control subjects (16 healthy and 15 pts after auto-HCT, 47 samples) were negative for MSI. MSI was observed in 37 (52%) allo-transplanted pts. MSI+ pts were significantly older than MSI-patients (median age 60y vs 48y, p<0.05). By using logistic regression analysis we found that the relative risk for MSI was 2-fold higher in pts who experienced extensive chronic GvHD as compared to pts with no GvHD. Although the median follow up in MSI+ pts was significantly lower than in MSI-(336 vs 669 days, p<0.05), secondary malignancy (5 skin-and 1 adeno-Ca but none in the oral cavity) was diagnosed in 5 (14%) of the MSI+ pts) and only in 1 (3%) MSI-pt (p<0.05). Other clinical features were not significantly different between MSI+ and MSI-pts. In an vitro mutation analysis model we tested the hypothesis that an alloantigenic stimulus is substantially involved in the mutation process. Briefly, keratinocyte (HacaT) cells were transfected with a plasmid vector carrying a neomycin selectable marker, a Hygromycin Resistance (HygR) sequence and a (CA)13 repeat. In this system, DNA slippage mutations become detectable after hygromycin treatment as HygR+ colonies. The mutant fraction was expressed as the number of HygR+ colonies corrected for relative cell survival. Untreated cells served as controls. Treatment of stably transfected HacaT cells with TNFa (25-100ng/ml, 24h), TGFb (5-20ng/ml, 24 h) and supernatant from a Mixed Lymphocyte Culture (MLC, 24h) didn't cause any detectable induction of genomic instability (GI). Treatment with H202 (20-40µÌ, 1-24 hours) resulted in a time and dose dependent GI induction (max 3.5 fold). Cocultivation of HacaT cells with stimulated lymphocytes from MLC resulted in a 3.9 fold induction of the mutant fraction. In conclusion, our in vivo and in vitro data indicate that "alloantigenic reactions may induce genomic instability in the allotransplanted pts which might predispose them to secondary neoplasia.

The EBMT risk score predicts outcome after allogeneic HSCT in all haematological disease categories and is independent of stem cell source or conditioning intensity A. Gratwohl*, M. Stern, J. Apperley, T. de Witte, J. Passweg, V. Rocha, A. Sureda, R. Brand, D. Niederwieser 

Information on factors associated with outcome after allogeneic HSCT is a prerequisite for risk adapted strategies.

Five key factors form the basis of the EBMT risk score: stage of the disease (early 0, intermediate 1, advanced 2), age of the patient (< 20 y 0, 20-40 y 1, > 40 y 2), time interval from diagnosis to transplant (< 1y 0, > 1 y 1), histocompatibility (HLA-id sibling 0, others 1) and donor recipient gender combination (other 0, female donor for male recipient 1). They were identified and validated in several independent series of CML patients, but not yet in other diseases. We examined 53140 patients, 34 y of age (median, 0-77y range), 58.7% male with an allogeneic HSCT for AML (15126; 28.6%), ALL (10756; 20.2%), CML 12321; 23.2%), MDS (4112; 7.7%), MPS (1184; 2.2%), Lymphoma (4165; 7.8%), Myeloma (1329; 2.5%) or SAA (4057; 7.6%) between 1980 and 2005. Donor was a HLA id sibling in 78.2%, other donor in 21.8%. Stem cell source was bone marrow in 64.6%, peripheral blood in 34.9% and cord blood in 0.5%.Conditioning was standard in 86.5%, reduced in 13.5%. Each risk factor was tested individually by multivariate analysis and confirmed as cumulative dose response risk in all subcategories with two exceptions: stage was not applicable in SAA, time interval was not applicable in patients in 1st CR. Cumulative incidence of transplant related mortality (TRM) at 5 years increased with the risk score from 15.4% (score 0, 3500 pts) to 22.1% (score 1, 10428 pts), 27.6% (score 2, 13677 pts), 32.4% (score 3, 11729 pts), 37.9% (score 4, 8159 pts), 42.6% (score 5, 4843 pts) and 49.7% (score 6/7, 1250 pts). After stratification by risk score, underlying disease had only a minor impact on the rate of TRM. Inside the risk score categories, TRM improved significantly during the period of observation (RR 1980 (RR -1989 RR 1990 RR -1999 RR since 2000 0.50 ). Absolute TRM rates declined less markedly (1980-89 36%; 1990-1999 31%; since 2000 27%) due to a shift towards higher risk patients in more recent years. The EBMT risk score separated risk categories in all diseases, for all donor types, for all stem cell sources and for patients with reduced or standard conditioning. These data show that risk categories for outcome after allogeneic HSCT can be defined. They can be integrated into risk assessment algorithms and form the basis for individualised risk adapted strategies when transplant and non transplant strategies are available as treatment options.

Working Party Solid tumours 114 Reduced-intensity allogeneic transplantation for breast cancer D. Blaise* (1) , A. Gonçalves (1) , S. Fürst (1) , J.O. Bay (2) , C. Faucher (1) , M. Michallet (3) , J.M. Boiron (4), J.Y. Cahn (5) , N. Gratecos (6) , M. Mohty (1) , C. Chabannon (1) , G. Gravis (1) , B. Esterni (1) , J.M. Extra (1) , P. Viens (1) (1)Institut Paoli-Calmettes (Marseille, FR) ; (2) Centre Jean Perrin (Clermont Ferrand, FR) ; (3)CHU Edouard Herriot (Lyon, FR) ; (4)CHU Haut Lévèque (Bordeaux, FR); (5)CHU Jean Minjoz (Besançon, FR); (6)CHU Cimiez (Nice, FR) We initially treated 18 pts with Allo SCT for advanced metastatic Breast Cancer. All pts (age: 45 (27-57)) underwent ASCT after the same reduced intensity conditioning (RIC) (Fludarabine (150mg/m2), Busulfan (8mg/kg) and Thymoglobulin (2,5mg/kg) or TLI (1 Cgy)) from a HLAidentical sibling (BM: 22%; PBSC: 78%) followed by CSA. Prior to ASCT a median of 3 lines of treatment (1-7) were administered over a period of 1452 days . Nine pts underwent autologous SCT at a median time of 1130 days (88-3012) prior to ASCT. All pts were measurable and had a median of 2 metastatic sites (1-4) (liver:72%, bone:50%, lung:22% and brain:11%): according to RECIST criteria, 13 (72%) and 5 (28%) pts had progressive (PD) and stable disease (SD) respectively. None of the 18 pts died from TRM. Two of them achieved partial remission (PR) at 60 and 150 days respectively (objective response (OR): 18% (0-36). All pts but 1 eventually progressed and died from disease (2 year overall survival (OS): 22%(9-45)). Results are dramatically different in regards to disease status at time of transplant. While outcome is uniformly poor for pts with PD, patients with SD achieved a 40% (0-80) OR rate for a 50% (19-80) OS at 2 years with 3 (60%) patients surviving more than 2 years (640+, 834-and 1246-), which is significantly different from pts with PD. We established that RIC-ASCT can be safely performed in BrC pts, whereas highly PD pts do not benefit from this approach. We run a second trial in less advanced disease to confirm encouraging results (present accrual: 15 pts). All 33 patients will be presented. However it seems that curability in BrC will be achieved only in pts in the initial disease phase: target population would need a careful selection on individual prognosis factors indicating their poor short term poor outcome: this represents the ultimate goal for future investigations. Supported in part by a grant from the French Ministry of Health (PHRC 2000; PHRC 2003 ) and a special grant (pole Areca) from the Association pour la recherché contre le Cancer (ARC)

We have utilized autografting to achieve maximum tumor reduction before proceeding to non-myeloablative allografting. This strategy could provide the benefit of a conventional allograft, but with reductions in the typical acute toxicities and associated mortality of myeloablative conditionings. Between September 1997 and April 2004, we enrolled 17 patients with metastatic breast carcinoma. Median age was 41 years. At the time of autografting, the patients had received a median of 3 (range, 2-5) previous chemotherapy lines; 14 patients had received hormone therapy, and seven patients had undergone radiotherapy on bone lesions. The primary endpoint of this study was the decrease of non-relapse mortality (NRM) from the current 20-35% noted after myeloablative allografting. Patients received autografting at a median of 53 months (range, 14-152) from the diagnosis of breast cancer. No patient died after transplant. One patient who had been in complete remission and two who had been in partial remission before autografting remained in complete or partial remission. No non-relapse mortality was noted in the first 100 days after non-myeloablative allografting. Thirteen patients achieved full chimerism. Five patients (29%) developed grade II-III acute GVHD, while six patients developed chronic GVHD (five patients with extensive disease) and needed intensive immunosuppressive therapy. We have recently reported a subsequent patient transplanted from her HLA-identical sister. Disappearance of liver, adrenal, mediastinal, pleural, and diffuse nodes and bone metastases, observed simultaneously with clinical chronic GVHC 5 months after non-myeloablative allografting, suggested a profound graft-versus-tumor effect.

Renal cell carcinoma (RCC) has recently been identified as being a target for GVT effect. Since 1999 there has been a number of publications describing GVT effects in patients with RCC undergoing mostly reduced intensity transplantation. At the NHLBI, patients have been conditioned with cyclophosphamide (60mg/kg x 2) and fludarabine (25mg/m² x 5) then transplanted with a G-CSF mobilized blood stem cell allograft from their HLA identical or single antigen mismatched related donor. Twenty-nine of 74 patients have had disease regression consistent with a GVT effect (39.2 % cumulative incidence of a complete response + partial response). A better understanding of the immune cells and their target antigens that mediate tumor regression could potentially lead to the development develop more effective HCT approaches for solid tumors. Recently, T-cells with in vitro tumor cytotoxicity patterns consistent with recognition of minor histocompatibility antigens and tumor restricted antigens have been identified in some responding patients. The identification of tumor restricted antigens targeted by donor immune cells could lead to the development of transplant approaches that enhance GVT effects while avoiding GVHD through tumor vaccination or the adoptive infusion of in vitro expanded donor T cells with tumor antigen specificity. We detected RCC-reactive CD8+ T-cells by ELISPOT analysis in the blood of several responding patients with metastatic RCC following HCT that were absent before transplantation. We successfully generated donor CD8+ T-cell clones from lymphocytes obtained from these patients that have direct cytotoxicity against the patient's RCC cells. In one responding patient, cytotoxic T-lymphocytes and T-cell clones with RCC-specific tumor cytotoxicity were isolated from the blood after transplantation. Utilizing cDNA expression cloning, we identified an HLA-A11-restricted 10-mer peptide (named CT-RCC-1) to be the target antigen of these RCC-specific Tcells (Takahashi Y, Harashima N. et al-J Clin Invest 2008 in press) CT-RCC-1-specific T-cells were detected by tetramer analysis in the patient's blood after tumor regression but not before HCT. Tetramer analysis of 8 HLA-A11+ RCC transplant recipients showed CT-RCC reactive T-cells expanded significantly in all 3 responders in contrast to the 5 non-responders, where only 1/5 showed an increase in CT-RCC-1 reactive T-cells.

The genes encoding the CT-RCC antigen were found to be derived from a human endogenous retrovirus (HERV)-E previously unknown to be expressed in any human tissues; this HERV-E was found to be expressed in the majority of RCC tumor lines and fresh RCC tissue but not in normal kidney cells or other normal tissues. This is the first solid tumor antigen identified using allogeneic T-cells from a patient undergoing HCT. These data suggest this HERV-E is transcriptionally active in RCC, encoding an immunogenic antigen that is over-expressed in RCC which could be a potential target for cellular immunity.

Update of the results of high-dose chemotherapy as primary or salvage therapy in germ cell tumours G. Rosti* (1) , U. De Giorgi (1) , P. Pedrazzoli (2) , M. Bregni (3) ( 

Cisplatin-containing regimens cure nearly 80% of patients with advanced germ cell tumors. HDCT has been extensively used in the last 20 years with somehow controversial results. The EBMT IT-94 study on relapsing good-risk patients has not shown any difference in overall survival comparing four courses of standard second line therapy versus one late intensification single shot approach, even if patients achieving CR did significantly better if randomized in the high-dose arm. A phase III US trial in patients with poor prognosis, treated upfront, even if not showing an advantage for HDCT, has shown a significant difference for those with unsatisfactory marker decline. An input of the possible role of HDCT in relapsing/refractory patients came from the retrospective data of the Indiana University of tandem HDCT with carboplatin and etoposide in a large series of consecutive men with metastatic testicular cancer that had progressed after receiving cisplatin-containing combination chemotherapy. This study shows 70% and 50% four-year disease-free survival in patients who received HDCT as second-line or third-line or later therapy, respectively. As it is a retrospective review, one may argue that the results are biased by patient selection. This does not seem to be the case, however, as even patients with very poor prognosis achieved long-term disease-free survival -50% of survivors were classified high-risk by the IGCCCG classification and 45% had platinum-refractory disease. It is important to note that all patients in this series received peripheral-blood progenitors as sources of hematopoietic stem cells. This strategy allowed a rapid engraftment, thereby permitting the administration of two courses of high-dose In addition, peripheral-blood progenitors were enriched for CD34+ hematopoietic cells, a procedure which may have a role in eliminating possible cancer cells from the graft. We believe that, on the basis of the robust data provided by Einhorn and colleagues, a well-designed randomized trial of HDCT versus conventional-dose chemotherapy should be performed in patients with poorprognostic clinical features who relapse after initial chemotherapy. At present, there should be no debate on the use of tandem-HDCT in patients with cisplatin-refractory germ-cell tumors and those who have failed second-line therapy. GITMO and IGG (Italian Group for Germ cell Cancer) are planning a network of centres in Italy to refer such patients for the tandem HDCT.

Autologous Stem cell Transplantation International Multiple Sclerosis trial (ASTIMS, Eudract number 2007-000064-24, supported by EBMT; www.astims.org) is now a multicenter, prospective randomized phase II study. The primary endpoint is the number of new T2 lesions on MRI. The investigational treatment comprises mobilization with CY and G-CSF and conditioning with BEAM followed by ASCT and ATG compared to 6 monthly i.v. pulses of mitoxantrone at 20 mg followed by 1gr of methylprednisolone. At the moment 18 patients have been enrolled (January 2008): 4 in Barcelona, 3 in Genova, 3 in Modena, 3 in Florence, 2 in Chieti, 2 in Reggio Calabria and 1 in Bergamo. The ASTIMS trial is therefore still going on, with the aim to arrive within 2 years from now at the new target of 30 enrolled cases. In the meantime, more than 400 are the MS treated in the world with AHSCT and a few phase I/II are running with the aim to identify the clinical characteristics of the patients who can really take advantage from the procedure or to evaluate the efficacy of low intensity conditioning regimens. The study of all the 58 cases treated in Italy in the last 10 years with the same regimen (BEAM and ATG), same inclusion criteria and followed by the same neurohematological teams involved in the prospective study supported by GITMO, showed that patients with a relapsing remitting clinical course respond significantly better than secondary progressive cases to AHSCT, indicating the population of patients who have to be selected in the future for the design of prospective studies with a clinical endpoint.

The ASTIS Trial J.M. van Laar* (1) , D. Farge (2) , A. Tyndall (3) , o. ASTIS investigators (4) (1) Newcastle University (Newcastle, UK); (2) Hopital St Louis (Paris, FR); (3)Basel University Hospital (Basel, CH); (4)JCUH (Middlesbrough, UK) Background: High dose immunosuppressive therapy (HDIT) and hematopoietic stem cell transplantation (HSCT) is a novel treatment for patients with severe systemic sclerosis (SSc) . Previous studies showed durable responses in two thirds of patients up to 7 yrs after HSCT (1) . This treatment modality is now further investigated through the ASTIS-trial (autologous stem cell transplantation international scleroderma trial), a prospective, controlled, randomized trial to compare safety and efficacy of HDIT + HSCT versus monthly i.v. cyclophosphamide in SSc patients at risk of major organ failure or early mortality. Objectives: To evaluate whether HDIT + HSCT is superior over conventional treatment in terms of safety and efficacy in SSc patients, and to assess potential predictive factors of response. Methods: SSc patients with early active diffuse disease with or without major organ involvement are eligible. SSc patients randomized to the transplant arm undergo mobilization with cyclophosphamide 2x2 g/m², conditioning with cyclophosphamide 200 mg/kg, rbATG 7.5 mg/kg, followed by reinfusion of CD34+ selected autologous HSCT. Those randomized to the control arm are treated with 12x monthly i.v. bolus cyclophosphamide 750 mg/m². The primary endpoint is event-free survival, defined as survival until death or development of major organ failure during 2 years follow-up.

Progression-free survival is the main secondary endpoint. Results: One hundred eleven SSc patients have been enrolled in 25 centers per January 2008: 43 male, 68 female, mean age 43 yrs, mean modified Rodnan skin score 26, mean disease duration 1,8 yr, mean VC 81%, mean DLCO 59%. Sixty-one patients were randomized to the transplant arm, 50 to the control arm. No unexpected toxicities have yet been observed in either arm with a median follow-up of 36 months (range . Grade 3,4 toxicities occurred in 15/43 transplant patients and in 13/48 controls (p=0.42). ATG-related toxicity led to its discontinuation in 12/35 transplant patients. Two fatalities in the transplant arm were categorised as probably treatment-related. Conclusion: The ongoing ASTIS trial has enrolled 111 patients sofar. Treatment-related mortality and number of patients with serious adverse events of stem cell transplantation are lower than previously reported in registry analyses. References: 1. van Laar JM, Farge D, Tyndall A, on behalf of the EBMT/EULAR Scleroderma Study Group. The ASTIS-trial, hope on the horizon. Ann Rheum Dis 2005;64:1515.

Standard NIH or Eurolupus cyclophosphamide (CY) protocols and mycophenolate Mofetil (MMF) as induction therapy in severe BILAG A SLE is still associated with 20 % failure, 50% relapse and 10% to 15 % death at 10 years In the absence of a single standard treatment worldwide for refractory SLE, phase I-II studies analysed the use of: a) rituximab (anti CD20 mAb) in more than 1 000 patients showing complete to partial early response around 100% with relapse in 50 to 60% of the cases; b) autologous Hematopoietic Stem Cell Transplantation (HSCT) since 1997 under the auspices of the joined EBMT-EULAR working party, reporting durable remission with reduced or no immunosuppressive drug requirement in 66%, one-third of whom later relapsed to some degree with a 74 ± 7% (n= 62/79) overall survival at 5 years for SLE among the 863 HSCT procedures registered: in 2007 in the EBMT data base. The North American, mostly single centre experience showed higher rates of remission with also some relapses. Maintenance immunosuppression after induction of remission may decrease the return of disease activity. This was the basis of the EBMT approved ASTIL trial: a prospective randomized open, multicenter, phase II b study to compare the efficacy of autologous HSCT with rituximab as remission induction, followed by MMF (2 g /day) as maintenance in both arms for severe SLE patients with disease duration ≤ 5 years since the diagnosis and sustained or relapsed active BILAG A SLE. This analysed describes the outcome of pediatric patients receiving hematopoietic stem cell transplantation (HSCT) to treat severe refractory autoimmune cytopenias. The registry of the EBMT contains data on 16 patients receiving 19 transplants. Patients had autoimmune haemolytic anemia (7), Evans's syndrome (7) , Immune thrombocytopenia (3), pure red cell aplasia (1) and autoimmune lymphroliferative synrome (1) . 15 patients were males with a median age at diagnosis of 4 years (range 0.3-16 years) and a median age at transplant of 7.8 years (2-17 years) . The median disease duration prior the transplant was 41 months (range 2-115 months) and all patients failed multiple prior treatments. Transplant were autologous for 7 and allogeneic for 12 patients, 6 of these transplanted from an HLA identical donor, 2 from a family mismatched donor and 4 from a matched unrelated donor. One patient received 2 autologous transplant while another patient received an allogeneic transplant because a relapse after the first autologous transplant . The stem cell source was mobilized PBSC in 1 transplant, bone marrow in 6 and cord blood in 2 patients. The graft was T depleted in 4 of 7 recipients of autotransplant and 3 of 12 allotransplant recipient. The conditioning regimen used were heterogeneous. 3 patients died of treatment related mortality, 2 in the allo and 1 in the autologous group for a TRM of 14 % Eights patients had a complete and continous response after the transplantation although 1 of these died for secondary malignancy. 3 patients relapsed after the procedure (1 in the allo and 2 in the autologous group) and one of these died for disease progression. 3 patients were not evaluable for response. The present analysis has some limitations because treatment protocols, mobilization and conditioning regimen were heterogenous and doesn't allows a detailed analysis of these factor moreover these preliminary data suggest that autologous and allogeneic HSCT may induce response in half of patients with severe autoimmune cytopenia of long duration unresponsive to several therapeutic options.

124b Long-term follow-up of autologous stem cell transplantation for juvenile idiopathic arthritis N.M. Wulffraat Wilhelmina Children's Hospital (Utrecht, NL) The majority of children with Juvenile Idiopathic Arthritis can nowadays be treated adequately. However despite the use of combinations of antirheumatic drugs, corticosteroids and the newer so called biologicals (blocking the TNF, Interleukin 1 or Interleukin 6 pathways) a proportion of children with arthritis remain resistant also to these therapies and suffer from a very severe, debilitating and potentially fatal disease. For such children autologous stem cell transplantation (ASCT) is successfully performed since 1997. Here we describe the long term outcome of the initial cohort of children with resistant Juvenile Idiopathic arthritis, treated with ASCT. The initial cohort of children was treated with a conditioning regimen containing Cyclophosphamide, anti thymocyte globulins and low dose Total Body irradiation. Overall favourable responses were seen, with a drug free remission rate of 50-55 %. In the more recent years late relapses were noted with lower percentages for drug free long term outcome. Special emphasis is given on 2 cases showing very late relapses, occurring after 7 and 9 years. The observed relapses are often less severe compared to the situation before SCT and can be treated successfully with conventional drugs in the majority of cases. More recently, ASCT was performed in 4 JIA children with a fludarabin containing regimen in stead of low dose TBI. With a 4 to 5year follow up, these 4 patients are all in drug free full remission. Allogeneic transplant with an HLA matched family donor was reported in 2 JIA cases. Follow up of 1 and 3 year is sofar show clinical disease remission and tapering of medition.

In conclusion, given the favourable long term outcome, SCT remains a valuable treatment option for children with drug resistant JIA. S7 126 Multipotent mesenchymal stromal cells in the treatment of autoimmune diseases A. Tyndall* (1) , F. Dazzi (2) 

Multipotent mesenchymal stromal cells ( MSC) isolated from the bone marrow and other sites are currently being studied to determine their potential role in the pathogenesis and/ or management of autoimmune diseases. In vitro studies have shown that they exhibit a dose dependent antiproliferative effect on T and B lymphocytes, dendritic cells, natural killer cells and various B cell tumour lines, an effect which is both cell contact and soluble factor dependent. These soluble factors include TGF beta, IL-10, indoleamine 2,3 dioxygenase, IL-1RA, and HLA-G among others. Anti proliferative and immunomodulatory mechanisms are probably multiple and most likely due to the induction of arrest of the cell cycle in G0/G1. A plethora of phenotypic definitions and experimental conditions accounts for some of the variation in in vitro phenomena being reported Previous assumptions that MSC are immunoprivileged have been challenged by recent animal data in non immunosuppressed hosts. Animal models of autoimmune disease and tissue injury (ischemic kidney, chemically induced lung fibrosis and liver toxicity) have mostly shown a positive clinical response, with some early warning signs in a melanoma model concerning tumour surveillance. A limited number of patients suffering from acute graft versus host disease have been treated with MSC as well as sporadic case reports and small uncontrolled series in multiple sclerosis and Crohns disease. Prospective phase I trials are starting in multiple sclerosis and Crohns disease and being considered in inflammatory rheumatic diseases. An international interdisciplinary data base has been developed to exploit the collective experience.

Sirolimus is associated with veno-occlusive disease of the liver after myeloablative transplantation C. Cutler*, K. Stevenson, H. Kim, P. Richardson, V. Ho, C. Revta, R. Ebert, D. Warren, J. Koreth, P. Armand, E. Alyea, R. Soiffer, J. Antin Dana-Farber Cancer Institute (Boston, US) Veno-occlusive disease of the liver (VOD) is an uncommon but important cause of mortality after allogeneic transplantation. To determine if use of the immunosuppressive mTOR inhibitor, sirolimus, is a risk factor for VOD, we performed a retrospective review of VOD incidence and risk factors at our institution since 2000, when we began using sirolimus. Methods: Review of electronic medical records of all transplant patients undergoing TBI-based myeloablative transplantation with adult stem cell donors was performed. Results: 510 patients transplanted between 1/2000 and 5/2007 were identified and stratified by sirolimus use (260 exposed, 250 unexposed). Sirolimus patients received sirolimus/tacrolimus ± methotrexate; all others received tacrolimus/methotrexate as GVHD prophylaxis. There were no differences in the age, gender, donor-recipient gender match or diagnoses between cohorts. Sirolimus patients were more likely to have unrelated or mismatched donors, were more likely to have received PBSC (p<0.001 for both) and were less likely to have Gr. II-IV acute GVHD (26 vs. 38%, p=0.004) in comparison with non-sirolimus patients. The incidence of VOD in the sirolimus group was 15% and was 6.4% in the unexposed (RR 2.3, p=0.003), but VOD occurred later among sirolimus patients (22 vs. 15 days, p=0.12) . Among MRD recipients, the RR was 2.6 (12.2 vs. 4.6%, p=0.06). When adjusted for age, gender match, stem cell source, HLA match (MRD vs. URD/mismatch), and transplant risk (standard vs. high), sirolimus use remained a significant risk for VOD (Adjusted OR 2.54, p=0.006). Cause-specific mortality related to VOD was similar in sirolimus and non-sirolimus patients. Despite the increase in VOD, treatment-related mortality was similar in all sirolimus and non-sirolimus patients and among MRD sirolimus and non-sirolimus patients at 1 year (19 vs. 21%, and 12 vs. 13%, both p=NS). In addition, there was a trend towards increased overall survival (OS) for all sirolimus patients (3 yr OS 54 vs. 49%, p=NS) and for MRD sirolimus patients (3 yr OS 67 vs. 53%, p=0.08). In a Cox regression model, age > 50 (p=0.001), donor match (p= 0.01) and VOD (p<0.001) but not sirolimus use (p=0.14) were significantly associated with overall survival. Conclusions: Sirolimus is associated with VOD after TBIbased myeloablative transplantation. Despite this association, transplant outcomes appear equivalent or better than standard tacrolimus/methotrexate based immunosuppression.

Physical health can be compromised in very long-term survivors after HSCT compared to their respective donors but not mental health: a paired analysis T. Daikeler, A. Rovo*, M. Stern, J. Halter, J.D. Studt, A. Buser, D. Heim, J. Rischewski, M. Medinger, A. Tyndall, A. Gratwohl, A. Tichelli University Hospital (Basel, CH) With the improvement of prognosis, health status and functional well being of long term survivors after HSCT become an important issue. We performed a cross-sectional prospective study on 44 long-term survivors and their respective sibling donors at a median follow up of 17.5 years (range 11-26) after HSCT. The median age of the recipients and donors at time of the study was 44.3 (24-63) and 43.4 years (22-61) respectively. Both recipients and donors were seen on the same day for evaluation. The short form-36® (SF-36) Health Survey, which provides a generic health status measurement through 36 items assessing 8 concepts of health was used. Three of the items measure physical health (PF, RP, BP), two measure both physical and mental health (GH, VT), and three measure mental health (SF, RE, MH). In addition there are two summary scores for physical (PCS) and mental (MSC) health. For statistical analysis norm-based scoring (NBS) was applied, where 50 is the mean score, and 10 the standard deviation of a defined general population. Paired analysis between donors and recipients were performed for detecting differences. All scored items of recipients as well as of the donors were within the range of one standard deviation of the norm-based population. All scores concerning physical well being except one, (RP), were statistically lower in the recipients than in their donors. In contrast, there was no difference in scores concerning mental well being. This is confirmed by the summary measurements of physical health (PCS) with 52,8 in the recipients and 57,1 in the donors , (p=0.001) and mental health (MCS) with 50,8 versus 52,9 (p=0.831) . Physical health (PCS) was lowest in patients with severe chronic GvHD compared to their donors (47,2 versus 57,2) (p=0.05), age older than 25 years at HSCT 50,7 versus 56,2 (p=0.024), older than 42 years at the last control 52,2 versus 55,63 (p=0.05) and for female patients 51,8 versus 57,7 (p=0.024) . None of the factors had a statistical impact on mental heath status (MCS, P>0.05).

In summary, quality of life of long term survivors after HSCT measured with the SF-36 questionnaire is still within the normal variation of the general population. However, when compared to their respective donors, the physical health status is significantly compromised in the recipients. Severe GvHD, older age and female gender are associated with an inferior physical health status.

A retrospective analysis of sexual function, fertility and endocrine status in male long-term survivors of allografts for haematological malignancy I.H. Gabriel*, R. Szydlo, M. Klammer, R. Patterson, N. Swan, N. Salooja, E. Olavarria, E. Kanfer, D. Marin, A. Rahemtulla, J.F. Apperely Imperial College Healthcare NHS Trust, Hammersmith Hospital (London, UK) Steady improvements in the outcome of allogeneic stem cell transplantation (allo-SCT) have resulted in significant numbers of long-term survivors and an increasing focus on factors impacting quality of life (QoL). Post transplant infertility and sexual dysfunction are two such factors. Using a questionnaire we audited 150 male survivors of allografting for haematological malignancies at our centre. 81 men (54%) responded. The median age at SCT was 37.5 yrs. The median time from transplant was 6.9 yrs and 97% were >3 years from SCT (63% >5 years). 76% had returned to full pre-SCT sexual activity, however, a number of problems in sexual function were reported. 49% complained of new persistent erectile dysfunction (ED) (normal prevalence 2-20%). ED affected men of all ages (40%, 47% and 51% at < 30 years, 30-40 years and >40 years respectively). ED was seen in 52% of recipients of TBI vs 22% of those who had not received TBI. 30% experienced penile glans dryness, previously unreported which appeared to be closely associated with chronic GvHD(p 0.019). In addition urethral constriction, phimosis, genitourinary infection and Peyronies's disease all complicated SCT. 48% of men reported reduced libido with 37% of ED patients reporting normal libido. 13% and 17% of men suffered premature or painful ejaculation respectively. 43% of male survivors described a negative impact of infertility on themselves or their partners and 9% had utilised assisted fertility/donor insemination. Other problems included dyspareunia, inability to use condoms, sicca syndrome, and poor body image. Factors forming barriers to new, or worsening existing, relationships were most prominent in younger patients. Compared to values pre-transplant FSH was raised in 100% by 6 months post-transplant and remained high at 5 yrs (p=0.003), indicating long-term damage to Sertoli cells. LH was significantly elevated at 12 months in 15% compared to baseline (p=0.005) and normalised by 5 yrs suggesting Leydig cell recovery. Testosterone levels were normal in 98% of men at 12 months. Lack of physician continuity, language, presence of visiting fellows/students and the impression that physicians are only concerned with managing malignancy were barriers discussing these issues at routine follow-up appointments. A high prevalence of sexual dysfunction exists post-SCT. Increased awareness of these complications and their effects on patients and their families would permit prompt and appropriate management.

Secondary malignancies in recipients of allografts for chronic myeloid leukaemia in chronic phase using HLAmatched sibling or volunteer donors I.H. Gabriel*, S. Avery, R. Szydlo, N. Salooja, E. Olavarria, E. Kanfer, M. Klammer, A. Rahemtulla, J. Goldman, J.F. Apperley Imperial College Healthcare NHS Trust, Hammersmith Hospital (London, UK) During the last decade imatinib has replaced allogeneic SCT as first line treatment for chronic myeloid leukaemia (CML) and second generation tyrosine kinase inhibitors now compete with allografting as second line therapy. However over the same period there have been steady improvements in the outcome of allografting and therapeutic choices are increasingly complex. Improved information regarding the long-term side effects of both chemotherapy and allografting might help inform decision-making. We now update the incidence of secondary malignancy in 481 consecutive patients of median age 33.9 years, who underwent allografts for CML in first chronic phase (291 from sibling and 184 from unrelated volunteer donors). The median follow-up is 13.9 yrs. All patients received cyclophosphamide with either total body irradiation (TBI) (n=474) or busulphan (n=7). GVHD prophylaxis was cyclosporin, methotrexate, ± T-cell depletion. The probability of developing a secondary malignancy post allo-SCT was determined by the method of cumulative incidence (CI). The overall incidence in this group was 5.95% with 27 patients developing 29 new tumours. 31% were diagnosed within 5 years, 62% within 15yrs, 72% within 20 yrs. Men and women were equally susceptible to secondary malignancies. The overall incidence of developing a new malignancy in our cohort is 5.95% which is higher than in an age-matched population. Involved sites included: skin (5 including 2 melanoma, 2 squamous cell and 1 basal cell carcinoma), breast (4), high grade B-lymphoma, (4), tongue (3), colo-rectal (2), cervix (2), osteosarcoma (2), testis (1), bladder (1), mucoepidermoid (1), oesophagus (1), lung (1), penile shaft (1), and MDS (1). 2 patients had 2 malignancies. Pre-BMT treatment consisted of hydroxyurea alone in 10, busulphan in 2 and interferon in 1 and some combination of these in 9 others. The CI of secondary neoplasia increased with time reaching 1.7%, 3.3%, 6.3% and 14.2% at 5, 10, 15,and 20 yrs respectively. 13 patients have died, 10 directly attributable to the secondary tumour. As all patients had the same genetic abnormality and received only minimal prior cytotoxic therapy, this increase in neoplasia is likely to be due to SCT. These data highlight the need for close follow up and screening.

Prospective evaluation of oxygenation index and noninvasive-ventilation in patients suffering from acute lung injury after allogeneic transplantation M. Wermke*, S. Schiemanck, G. Höffken, G. Ehninger, M. Bornhäuser, T. Illmer University Hospital (Dresden, DE) Objectives: Respiratory dysfunction is a major cause of allogeneic transplant-related-mortality (TRM). Little is known about how to recognize and treat Acute Lung Injury (ALI) in this setting. We present the first prospective randomized clinical trial evaluating the use of Non-Invasive-Ventilation (NIV) for treatment of ALI in allogeneic transplantation. Methods: All patients (n= 530) undergoing allogeneic transplantation at a single center were investigated from 2001 to 2005. Oxygenation-Index (PaO2/FiO2) was monitored twice daily from the beginning of the conditioning regimen. Patients meeting criteria for ALI were randomized to receive either oxygen or intermittent NIV with positive end-expiratory pressure. Patients not responding to assigned therapy were allowed either to switch from oxygen to NIV or proceeded with treatment on intensive care unit (ICU). Results: Of 90 eligible individuals, 44 were randomized to receive oxygen and 42 to NIV, 4 patients withdrew consent. Oxygen-, NIV-and control-(440 patients without ALI) group were well balanced regarding known risk factors for TRM. Only TBI was significantly more frequent in patients with ALI. Patients with ALI showed significantly shortened short-and long-term overall survival (OS) when compared to controls (100-day-OS 65 vs. 85%; Median-OS 7 vs. 22 months). Oxygenation index was not only a major adverse prognostic factor; it also suggested ALI long before clinical parameters raised suspicion. Of 42 patients randomized to NIV only 10 (24%) did not respond and had to be transferred to ICU. In contrast 18 of 44 patients (41%) assigned to oxygen did not improve. Of these, 17 switched to NIV but only 12 patients had to be transferred to ICU. Thereafter, there was no significant difference between both groups in need for intubation (oxygen: 11 patients, NIV: 7 patients). NIV did not lead to improved survival in our study (100-day-OS NIV: 61%, oxygen: 68%; Median-OS NIV: 6 months, oxygen: 7 months). Survival of intubated patients was poor, as only 1 of 18 intubated patients survived for more than 100 days. Conclusion: Oxygenation-Index is an easily measurable early indicator of ALI and poor survival in allogeneic transplantation. NIV seems to reduce the need for ICU and consecutive intubation in patients not responding to oxygen. We were not able to show significant survival benefits for NIV-patients, presumably because patients not responding to oxygen were allowed to switch to NIV.

What to do when the first allogeneic stem cell transplantation fails? M. Kedmi, I.B. Resnick, B. Gesundeheidt , L. Drey, S. Samuel, R. Or, M.Y. Shapira* Hadassah -Hebrew University Medical Center (Jerusalem, IL) The failure of allogeneic stem cell transplant (allo-SCT) is usually cumbersome, we have retrospectively evaluated our experience in a 2nd allo-SCT. Patients: Out of 1533 allo-transplantees, 145 patients (93 males, 52 females, median age 20.7y(8m-68.4y)) underwent 2 or more allo-SCT. The indications for the 1st transplant were acute leukemia (93), chronic leukemia (17), lymphoma (3), other malignancies (3) and non malignant (29). The 1st to 2nd interval was 18d to 13.25y (median 98d). The most frequent indications for the 2nd SCT were basic disease (45), rejection (23) and engraftment failure (15). The 2nd SCT conditioning was radiation based (59)or chemotherapy based in the others and was myeloablative or reduced intensity (RIC) in 64 and 81 respectively. In 89 of the SCTs the original donor was used. 2nd donor matching was full in 92 transplants (family-84, unrelated-8) or mismatched in 53 transplants (51 and 2, family and unrelated donor respectively). Graft source for the 2nd SCT was BM (65) and PBSC (80). 38 of the grafts were T cell depleted. GVHD prophylaxis was given in 31 2nd procedures. The median survival from 2nd SCT was 70d. Despite the low rate of GVHD prophylaxis used only 51 and 16 of the patients developed aGVHD and cGVHD respectively. 29/145 patients (20%) transplanted survived a year after the 2nd SCT (figure 1). TRM was 66% including sepsis, liver and renal failure, neurologic toxicity, rejection and GVHD. Factors indicating higher chance for survival were non malignant disease, longer time between procedures (more then 1y), HLA matching and the use of RIC (figure 2). Age at transplantation, the indication for transplantation (relapse vs. others), the development of acute GVHD, radiation based vs. chemotherapy based conditioning, GVHD prophylaxis (either pharmacological or T cell depletion) or graft source were not shown to indicate better or worse prognosis. With a median follow up of 4.5y, 25 patients (17.2%) are still alive, out of which 18 are disease free. Conclusion: although highly toxic, a 2nd allo-SCT may be beneficial to some patients and lead to long term survival. It seems that the patients whom will gain the most out of 2nd allo-SCT are those who had a longer period between SCTs, have a matched donor or have non malignant diseases although 14 patients with malignant diseases (some with refractory disease) survived at least a year from the 2nd procedure while in CR. It is also noteworthy that the use of RIC improved outcome. S10 O136 Treatment of donor graft failure with autologous or allogeneic stem cell boost or a second allogeneic transplantation based on chimerism testing A. Shimoni*, N. Shem-Tov, A. Rand, E. Ribakovsky, R. Yerushalmi, I. Hardan, A. Nagler Chaim Sheba Medical Center (Tel-Hashomer, IL) Donor graft failure (GF) is a life-threatening complication of allogeneic stem cell transplantation (SCT). We performed this analysis to determine the rate and outcome of GF in the era of modern SCT. We retrospectively reviewed data of 491 SCTs from HLA-matched siblings (n=284), matched-unrelated (MUD, n=180) or alternative donors (mismatched-related, haplo-identical and cord-blood, n=27) performed in a single institution since 1/2001. GF was diagnosed in 25 patients (pts), cumulative incidence (CI) 5.2% (95%ci 3.5-7.6). GF was determined when ANC had not reached 0.5 x 10*9/L by day 21 (primary GF, n=21) or when ANC decreased irreversibly after engraftment (secondary GF, n=4). CI of GF was 2.5%, 6.8% and 23.4% after SCT from siblings, MUD or alternative donors, respectively (p<0.001) but was similar following myeloablative or reduced-intensity conditioning (5.7% and 4.6%, respectively). Pts with a predominant donor population in chimerism testing were given donor cell boost with no additional conditioning (n=10). Pts with a predominant host population were given autologous back-up cells (n=8) or a second SCT from a different donor (sibling-1, haplo-3, MUD-1) with nonmyeloablative conditioning. 18 pts survived > 1 week after second graft infusion and are evaluable for engraftment. 16 pts engrafted within a median of 10 days (range, 5-15). The probability and pace of engraftment was similar in the different approaches. 11 pts (44%) were able to be discharged home and 14 died; 2 early after diagnosis of GF with no intervention, 5 within one week of second graft infusion and prior to engraftment, 2 with no engraftment and 5 early after engraftment from infection (n=3), organ failure (n=1) and GVHD (n=1). With a median follow-up of 19 months (range, 3-68), 6 are alive and 5 additional pts died (relapse-3, GVHD-1, infection-1). The projected 2-year survival for all pts was 23% (95%ci 5-41). Interestingly, 4 pts given autologous cells had donor cell recovery, 1 had spontaneous autologous reconstitution within 3 weeks, 2 died within 2 months (GVHD-1, infection-1) with persistent donor cells and 1 remained complete donor until she relapsed 2 years later. In conclusion, treatment of GF with a chimerism directed method can salvage a subset of pts with GF. Reserving autologous and/or donor backup cells or an alternative donor is advisable in pts at high-risk of GF. The observation of allogeneic recovery after autologous boost is intriguing and of unknown mechanism.

At home autologous stem cell transplantation for haematological malignancies: the role of preparative regimens F. Fernández-Avilés*, M. Rovira, C. Martínez, A. Gaya, C. Gallego, A. Hernando, S. Segura, L. García, J. Güell, M. Valverde, E. Carreras, E. Montserrat Hospital Clínic (Barcelona, ES) Background: The aim of this study was to investigate the impact of the most commonly used regimens on the engraftment, toxicity and readmission rate after autologous stem cell transplantation (ASCT) in patients managed at home. Patients and Methods: At home ASCT (since day +1) was offered to all patients with a good performance status, a travelling time to the hospital of less than 60 minutes, and a caregiver available 24 h per day. In all patients the preparative regimen was administered at the hospital. Patients with lymphoma were treated with intensified BEAC or BEAM (mg/m 2 ) (BCNU 300, etoposide 1600, cytarabine 800, and cyclophosphamide (CY) 6000 or melphalan 140), patients with myeloma received melphalan 200 mg/m 2 and patients with leukaemia total body irradiation (TBI) 12 Gy and CY 120 mg/kg. All patients received the same supportive care, including prophylactic i.v. ceftriaxone once daily. Indications for re-admission to the hospital were: patient's or caregiver's desire; uncontrolled nausea, vomiting or diarrhea; mucositis requiring total parenteral nutrition or i.v. morphics; persistent fever; hemodynamic instability, pneumonia, or cardiac and/or respiratory distress. Results: Seventy-five patients were included in this study. Forty-five received BEAC (n=10) or BEAM (n=35) (n=45, group A), 19 melphalan (group B) and 11 TBI-CY (group C). Recovery (days) of granulocyte count above 0.5x109/l (group A: 11 (9-22), B: 12 (10-26) and C: 12 (10-22)) was significantly faster for patients from group A (A vs. B, p=0.02; A vs. C, p=0.05). Fever occurred in 87%, 47% and 73% of patients in the groups A, B and C, respectively (A vs. B, p=0.003). The median (range) days with fever were 2 (1-11), 1 (1-4) and 1 (1) (2) (3) (4) (5) for A, B and C groups, respectively (A vs. B, p=0.01; A vs. C, p=0.05 and WHO mucositis grade upper to 2 was observed in 40%, 5% and 27% (A vs. B, p=0.006). In group A, 9 (20%) patients needed re-admission by pneumonia (n=4) and persistent fever (n=5), as compared to only 1 (3.3%) because of persistent fever in the rest of patients (p=0.04). Conclusions: Despite a faster granulocyte recovery, patients managed at home after BEAC-BEAM presented a higher incidence and duration of febrile neutropenia, severe mucositis, and high rate of hospital re-admission, especially when compared with patients receiving melphalan. Based on this information we have established a more frequent and strict follow-up of patients having received BEAC-BEAM and managed at home.

Outcome of allogeneic stem cell transplantation in first remission for Philadelphia chromosome-positive acute lymphoblastic leukaemia following three schedules of imatinib-based chemotherapy B. Wassmann, N. Goekbuget, H. Pfeifer, D.W. Beelen, J. Dengler, N. Kröger, M. Stelljes, K. Kolbe, W. Bethge, M. Bornhäuser, H.-J. Kolb, M. Lübbert, M. Stadler, H. Serve, D. Hoelzer, O.G 

Background: Allogeneic SCT in CR1 is considered the best curative treatment option in Philadelphia chromosome-positive acute lymphoblastic leukaemia (Ph+ALL) pts., but relapse and transplant related mortality (TRM) remain significant limitations. Although Imatinib (IM) is considered standard element of front-line therapy, the impact of different treatment schedules is not known. Objective: To evaluate overall survival (OS), TRM and relapse risk (RR) in Ph+ALL pts. following allogeneic SCT in CR1 (n=108) after three different IM-based chemotherapy regimens. Patients: In a prospective, multicenter GMALL study, three successive pt. cohorts received IM according to one of three schedules: In cohort A (n=43) IM was administered alternating with chemotherapy beginning immediately after induction phase II (IPII)(med. age 44(23-62)yrs.). In cohort B (n=72)IM was started after completion of induction phase I (IPI) and given concomitantly with chemotherapy throughout IPII, 1st consolidation (cons.I) and up to yrs.). Pts. in cohort C (n=41) received front-line IM starting after a 5 day prephase and continued parallel with IP I+II, cons.I up to yrs.). Results: The proportion of pts. transplanted in CR1 was 33(77%), 49(68%) and 26 (64%) in cohorts A, B and C, respectively. Median follow-up since allogeneic SCT was 11 (0.8-68) , 24 (0.3-51)and 9 (1-22) mo.. Median OS was 11 mo., 37 mo. and not reached. Estimated OS at 22 and 48 mo. was 39%, 57% and 71% and 33% and 47%, respectively. TRM (35/108; 32%) was mainly due to infections (n=12; 34%) or GvHD (n=11; 31%). 29/35 (82%) TRM deaths occurred in pts. aged >35 yrs.. TRM at 3 and 12 mo. in pts. aged ≤ 35 yrs.(n=32) was 7% and 17% compared with 21% and 36% in pts. aged >35 yrs.(n=76)(p=0.05). RR was significantly lower in pts. with low or undetectable pretransplant Bcr-Abl levels (33% at 66 mo. vs. 100% at 18 mo., p=0.0002). Conclusion: All tested schedules enable allogeneic SCT in a high proportion of pts.. There was a trend towards superior OS in cohort C with no significant difference in TRM in the 3 cohorts. Age and pretransplant MRD levels had the greatest impact on treatment outcome: TRM was significantly higher in pts. >35 yrs. Both low and negative pretransplant Bcr-Abl levels were predictive of a low probability of relapse, whereas high levels were associated with a 60% relapse incidence within the 1st year. The impact of posttransplant IM on relapse risk remains to be determined.

Impact of FLT3-ITD on the outcome of HLA-identical stem cell transplant for adult AML with normal cytogenetics: substantial probability of leukaemia-free survival despite increased relapse risk. A retrospective analysis of the EBMT-ALWP S. Brunet *, M. Labopin, A. Gratwohl, A. Buzyn, J. Harousseau, J. Jouet, G. Socié, A. Rambaldi, M. Mothy, G. Cook, J. Sierra, V 

The prognosis of patients with acute myeloid leukemia (AML) and FLT3 internal tandem duplication (FLT3-ITD) is poor. It is unclear whether this mutation has impact on the outcome of allogeneic stem cell transplantation (Allo-SCT) for AML in patients with normal cytogenetics (NC). Different cooperative groups have obtained controversial results. For this reason, registry data including large numbers of patients are of interest. We analysed the predictive value of FLT3-ITD on relapse incidence (RI), non-relapse mortality (NRM) and leukemia-free survival (LFS) in patients with AML in 1st CR with NC who underwent a myeloablative allo-SCT from HLA-identical siblings and were reported to the EBMT. The series included 131 patients, 89 negative ITD/FLT3 and 42 (32%) FLT3/ITD positive reported between 2001 and 2007 with a median follow up time of 16 and 10 months, respectively. No significant differences were observed between the two groups regarding gender, median age (38 vs 43 yrs), FAB classification, number of induction courses to achieve CR, interval diagnosis to transplant, conditioning regimen, in vitro T-cell depletion, female donor to male recipient and CMV serostatus. In contrast, patients with FLT3-ITD+ had higher leukocyte counts (WBC) at diagnosis (16 vs 67.9 x10 9 /L, p=0.001) and more frequently peripheral blood was the source of stem cells (43% vs 67%, respectively). Univariate (table1) and multivariate analyses demonstrated the adverse impact of FLT3-ITD for LFS (HR 0.29, p=0.01) and relapse incidence (HR=3.17, p=0.008). It is remarkable, however, 59% of FLT3/ITD positive patients transplanted for AML in first CR remain alive and disease-free at 2 years. Other independent prognostic factors for LFS and relapse were: more than one induction course to achieve first CR. For relapse, higher WBC (HR=2.9, p=0.05) was also a significant prognostic factor. In summary, this analysis shows an adverse impact of FLT3-ITD on transplantation outcome. Despite this finding 59% of FLT3/ITD positive AML patients in first CR and normal karyotipe remain disease-free at 2 years, a proportion that seems higher than with other treatment options. It is becoming clear that a higher lymphocyte count one month after allogeneic stem cell transplantation (SCT) is associated with better transplant outcome in patients transplanted from an HLA-identical sibling. However, a predictive role of the day 30 post-transplant absolute lymphocyte count (LC30) in unrelated transplants is not defined. We studied the relationship between lymphocyte counts and other engraftment parameters on outcome in 102 patients with myeloid leukemia (54 acute myeloid leukemia, 38 chronic myeloid leukemia, and 10 myelodysplastic syndrome) receiving myeloablative SCT from an HLA-A, -B and -DR matched unrelated donor at Karolinska University Hospital, Stockholm from 1996-2006. Median recipient age was 37 years (range 0.5-58), 22 patients (22%) were under 18 years. Conditioning consisted of cyclophosphamide with Busulphan (n=61) or total body irradiation (n=41). Bone marrow (BM) was given to 44 patients and mobilized unmanipulated peripheral blood stem cells (PBSC) to 58. Median stem cell dose was 6.8 x 10 6 /kg (0. 2-56.4 ). Sixty-three patients (62%) received G-CSF post-graft. Immunosuppression used post graft was cyclosporine with four doses of methotrexate in 97 patients and other treatments in 5. Overall survival at 5 years was 61% and relapse-free survival was 56%. The incidence of aGvHD grades II-IV was 62% in patients with an LC30 of <0.2x10 9 /l, 33% if the LC30 was 0.2-1.0 x10 9 /l and 25% in patients with an LC30 >1.0 x10 9 /l (p=0.008). Transplant related mortality (TRM) was 34% in patients with an LC30 <0.2x10 9 versus 19% (LC30 of 0.2-1.0 x10 9 ) and 0% (LC30 >1.0 x10 9 )(p<0.001). Survival was significantly higher in 17 patients with an LC30 >1.0x10 9 /l, compared to 67 patients with an LC30 0.2-1.0 x10 9 /l, and 18 patients with <0.2x10 9 /l (91% vs. 60%, vs. 36% p=0.02 and 0.001 respectively). When analyzed as a continuous variable in multivariate analysis, a higher LC30 was associated with a lower incidence of acute GvHD grades II-IV, improved survival, less relapse and higher relapse-free survival (see figure) . These results indicate that the LC30 is a robust prognostic factor for transplant outcome in matched unrelated as well as matched related SCT for myeloid malignancies receiving either BM or PBSC with or without irradiation conditioning. Further research to identify the transplant conditions leading to prompt lymphocyte recovery might lead to global improvements in SCT outcome in unrelated SCT. S12 O141 GITMO survey on the outcome of 2333 acute myeloid leukemia patients receiving autologous stem cell transplantation in the old and recent era: final results of the multivariate analysis A. Olivieri*, B. Bruno, G. Meloni, M. Falda, A. Rambaldi, W. Arcese, E. Alessandrino, R. Scime', R. Lemoli, M. Cimminiello, A. Bacigalupo, A. Bosi on behalf of Gruppo Italiano Trapianto Midollo Osseo (GITMO) GITMO Registry data files from 2333 acute myeloid leukemia (AML) adults, autotransplanted from 1985 to 2004, have been evaluated to assess the outcome according to age, conditioning and Stem Cell source; 2032 patients received one autologous stem cell transplantation (ASCT); 94 two or more ASCT; 207 allogeneic transplant after failure of ASCT. Patients were categorized in 2 cohorts basing on the transplant era (before or after 1998); the two cohorts were significantly different as regard: age distribution (33% pts older >55 yrs transplanted after 1998 vs 10% pts >55 yrs before 1998); less pts <30 yrs received ASCT after 1998 (13% versus 28%); preparative regimens: TBI or BU-Cy were less common in the recent era; status of disease: 92% received ASCT in 1rst or 2nd CR before 1998, compared to 90% after 1998; stem cell Source: more pts (75%) received PBSC after 1998 versus 16% before 1998. With a minimum 8 yrs followup, OS did not differ in the two cohorts: 37% in the 1111 patients autotransplanted after 1998 vs 36% in the 1220 autotransplanted before 1998. As a preliminary analysis suggested a possible role of PBSC in reducing the overall NRM after 1998 (from 14% to 11% after 1998), we made a multivariate analysis to evaluate the main variables (Age, Stem cell source, Transplant Era, Disease status at transplant and type of conditioning) influencing the outcome, in terms of OS, NRM, DFS and Relapse Incidence. Age>55 yrs significantly worsened NRM (1,76 HR), OS (1, 76) and DFS (1,5 HR) regardless the transplant era; ASCT performed in advanced disease (>2nd CR) was associated with a 2,9 times increased risk of Relapse and NRM. Stem cell source did not significantly influence OS and NRM, but BM source, instead of PBSC, was associated with significantly reduced risk (0,79 HR; p=0.012) of relapse. Finally TBI regimens were associated with increased NRM compared to BU-CY (1,6 HR). Some conclusions can be drawn from this survey: 1-the advanced age (>55 yrs) remains an adverse factor both for NRM, OS and for LFS; 2-these data definitely show that PBSC are associated with increased risk of relapse after ASCT, without major impact on OS; 3-regimens including TBI are not recommended being associated with an increased NRM without a significant reduction of Relapse. The main efforts in the future should be aimed to reduce the relapse incidence probably by designing new conditioning regimens (and by targeting the minimal residual disease post-transplant) and by a cautious use of PBSC in patients receiving a short consolidation.

Up-front allogeneic stem cell transplantation as part of induction therapy in newly-diagnosed high-risk acute myeloid leukaemia -an update of a prospective phase II trial U. Platzbecker*, M. Füssel, M. Schaich, T. Illmer, B. Mohr, J. Schetelig, A. Kiani, C. Theuser, C. Thiede, G. Ehninger, M. Bornhäuser University Hospital (Dresden, DE) Poor-risk cytogenetic aberrations, bad response to the first cycle of induction chemotherapy (IC) or the presence of an FLT-3 receptor mutation define high-risk AML (HR-AML) and result in an increased risk of failure of long-term disease control. As a matter of fact, only a minority of this patient group proceed to HSCT due to treatment failure or death from infectious complications during IC. Therefore, there is substantial need to improve the outcome of these patients with HR-AML. We report results of an ongoing prospective trial evaluating an "early" HSCT applied during induction-chemotherapy induced aplasia in forty (n=40) newly-diagnosed HR-AML patients with a median age of 50 years (17-68). A median of 12 days (range 6-34) after the first (n=18) or second (n=22) cycle of IC patients received a reduced-intensity regimen that was based on fludarabine combined with either busulfan (n=4) or melphalan (n=36) followed by allogeneic G-CSF mobilized peripheral blood stem cells (PBSC) from related (n=12) or unrelated (n=28) donors. Twenty-six patients were not in complete remission before conditioning therapy was started with a median marrow blast count of 20 % (range 6-85). Patients with unrelated grafts received antithymocyte globulin and GvHD prophylaxis was performed with cyclosporine A only in all patients. All patients engrafted and went into remission. Acute GVHD grade II-IV occurred in 35 % and extensive chronic GvHD in 30% of patients. With a median follow-up of 17 months (range 1-91) the probability of overall and disease-free survival at 24 months is 68%. Early allogeneic HSCT as part of primary induction therapy seems to be an effective strategy in high-risk AML patients.

Comparison of up-front versus minimal residual disease triggered imatinib after stem cell transplantation for Philadelphia chromosome-positive acute lymphoblastic leukaemia: interim results of a randomised phase III study B. Wassmann*, H. Pfeifer, W. Bethge, M. Bornhäuser, J. Dengler, M. Stadler, D. Beelen, N. Basara, R. Schwerdtfeger, K. Schäfer-Eckart, L. Uharek, H. Serve, D. Hoelzer, O.G. Ottmann on behalf of the GMALL study group Background: Detection of minimal residual disease (MRD) following allogeneic SCT for Philadelphia chromosomepositive acute lymphoblastic leukemia (Ph+ALL) is highly predictive of evolving relapse. We have previously shown that only about 50% of pts. with Ph+ALL who convert to MRD positivity after allogeneic SCT achieve renewed PCR negativity in response to interventional Imatinib (IM), started a median of 4 mo. after SCT. Failure to rapidly achieve a complete molecular response was almost invariably associated with hematologic relapse. We hypothesized that earlier initiation of IM in the setting of a lower leukemic cell burden would increase response rates and improve DFS. Objective: To determine whether the earliest possible initiation of IM after SCT is superior to delayed IM triggered by reappearance of bcr-abl transcripts with respect to feasibility, tolerability and duration of molecular and hematologic remission. Patients: Six wks. after allogeneic SCT pts. were randomly assigned to receive IM either up-front (cohort 1)(n=17) or subsequent to detection of bcr-abl positivity as determined by real time quantitative and/or nested rtPCR (cohort 2)(n=17). Target dose of IM was 600 mg, but a lower dose of 400 mg was permitted if deemed necessary because of tolerability. IM was scheduled for a total duration of one year of PCR negativity. Results: To date, 34 patients have been enrolled. Of the 17 pts. randomized to each cohort, 15 each underwent SCT in first complete remission (CR1), 2 each in CR2. IM was started in 13/17 pts. in the up-front IM and 8/17 in the MRD-triggered cohort, with most pts. receiving 400 mg IM (10/13 pts. and 5/8 pts., respectively). Med. time from SCT to start of IM was 45 (cohort A) and 82 days (cohort B). After a med. follow-up of 260 (29-985) and 281 (31-1016) days in cohorts A and B, respectively, none of the 30 pts. transplanted in CR1 and 1 of 4 with SCT in CR2 relapsed. Three pts. (cohort A) died in CR, only 1 one of whom had actually received IM. IM was discontinued prematurely in 5/13 pts. in the IM up-front arm and 3/8 in the MRD-triggered IM cohort, due mostly to gastrointestinal toxicity and GvHD. Conclusions: With rigorous monitoring of MRD, both schedules of post-SCT IM (up-front versus MRD-triggered) are associated with a remarkably low relapse rate, although follow-up is still short. Tolerability of IM is poorer than generally experienced in non-transplanted patients. The routine use of IM after SCT is a promising strategy to improve outcome of pts. with Ph+ALL. We compared reduced intensity conditioning (RIC, n=488) with myeloablative conditioning (MAC, n=1596) in patients with AML undergoing hematopoietic stem cell transplantation (HSCT), using HLA-A, -B and -DRb1 identical unrelated donors, transplanted between 1999 to 2005 and reported to the EBMT. In the RIC group, median age was higher, transplant was performed more recently, time from diagnosis to transplant was longer, T-cell depletion was less commonly used, and more patients received PBSC vs. bone marrow (p<0.0001). In patients below 50 years of age in CR1, transplant-related mortality (TRM), relapse and leukemia-free survival (LFS) were similar in the two groups. In patients in CR2-3 below 50 years of age, 2-year probability of relapse was 31% in the MAC group, compared to 51% in the RIC patients (p=0.006). In these patients, TRM was 36% vs. 26%, and LFS was 44% vs. 36% at two years in the two groups, respectively (ns). In patients with AML advanced disease below 50 years of age, TRM, relapse and LFS did not differ significantly between RIC and MAC patients. At two years, LFS was 22% with MAC vs. 12% using RIC (p=0.19). In patients above 50 years of age in CR1 and advanced disease, relapse was not statistically different in the two groups. In CR2-3, MAC patients above 50 years had a probability of relapse of 18% vs. 47% for RIC patients (p=0.009). There were no statistically significant differences in TRM or LFS, using MAC or RIC in patients above 50 years. To conclude, using RIC as an alternative to MAC in patients with AML undergoing transplants with matched unrelated donors resulted in: no significant difference in TRM in patients above and below 50 years of age, significantly increased risk of relapse in patients treated with RIC in CR2-3. LFS was similar using the two conditioning regimens.

Higher incidence of relapse with peripheral blood as source of stem cells in adult patients with acute myelocytic leukaemia autografted in first remission N.C. Gorin (1) In the past 30 years the modalities of autologous stem cell transplantation (ASCT) in first remission (CR1) for patients with acute myelocytic leukaemia (AML) have evolved: The age limit for ASCT has been gradually extended up to 70 years . Total body irradiation (TBI) pretransplant has declined in favour of chemotherapy (CT). Leukapheresis products (PB) have replaced bone marrow (BM) as source of stem cells. We were interested in evaluating the potential impact of these modifications on the outcome post ASCT. A total of 7648 ASCT (2947 BM, 4701 PB) were reported from January 1985 to December 2006. 79% of all PB transplants were done after 1994. Therefore, we compared BM versus PB in patients patients transplanted after 1994: 1226 patients received BM and 4605 PB. The median follow up in the two groups were 46 and 16 (1-161) months respectively. Patients receiving PB were older (48y vs 42,p<0.0001), had more AML of the M5,6,7 categories (p<0.0001) and received less TBI (15% vs 30%, p<0.0001). By multivariate analyses, age was a significant factor with lower TRM, higher relapse incidence (RI) and lower LFS above 45 years. Failure to reach CR within 40 days was associated with a higher (RI) and a lower LFS. The use of PB as compared to BM significantly resulted in a higher RI (51 ± 1% vs 43±2%, p=0.003), and a lower LFS (44±1% vs 50±2%,p=0.04) with only a trend for a slight reduction in TRM (9±1% vs 12±1%, p=0.1 ). TBI was not a significant factor for outcome. We finally focused on good risk patients (CR within the first 40 days): The population consisted of 844 patients grafted with PB and 326 with BM. Patients receiving PB were older (p<0.0001), and received less TBI ( p<0.0001). Again in this good risk population the RI was significantly higher with PB (48 ± 2% vs 35±3%, p<0.01), with a trend for a worse LFS (47±2% vs 57±3%,p=0.1). We conclude that the shift to PB as a source of stem cells in the past 15 years has resulted in increasing the relapse incidence possibly through mobilisation of leukemic cells and /or insufficient purging of the autograft. S14 Experimental stem cell transplantation/ Stem cell research O146 Co-transplantation of placental derived mesenchymal stromal cells produces superior engraftment of umbilical cord blood compared to double unit umbilical cord blood transplantation S. Hiwase, P. Dyson, S. Young, B. To, I. Lewis* IMVS (Adelaide, AU) Double-unit umbilical cord blood transplantation (UCBT) has been shown to overcome some of the limitations of UCBT, particularly in adult recipients. Whether this simply reflects a cell dose effect has not been established. Co-transplantation of mesenchymal stromal cells (MSCs) has also been suggested as a means of enhancing engraftment and may be appropriate in patients where two suitably matched cords cannot be identified. In this study we have directly compared engraftment rates of double-unit UCBT with MSC cotransplantation. MSCs were obtained from placental tissue by enzymatic digestion and isolated by plastic adherence. Placental MSCs demonstrated fibroblastic morphology, immunophenotype and differentiation potential similar to bone marrow derived MSCs. In a NOD/SCID mouse model 4 groups of mice were compared: Group 1 received 5 x 10 4 CD34+ cells from a single cord unit (U1); Group 2 received 5 x 10 4 CD34+ cells from U1 + 4 x 10 4 MSCs; Group 3 received 2.5 x 10 4 CD34+ cells from U1 + 2.5 x 10 4 CD34+ cells from U2; Group 4 received 2.5 x 10 4 CD34+ cells from U1 + 2.5 x 10 4 CD34+ cells from U2 + 4 x 10 4 MSCs. In 4 independent experiments mean engraftment rates were: Group 1 28%, Group 2 47%, Group 3 24%, and Group 4 44%. Hence MSC co-transplantation produced superior engraftment when compared with either single unit (p=0.05) or double unit transplantation (p=0.04). Combining results demonstrated the superiority of MSC co-transplantation with mice receiving MSCs showing mean engraftment of 45% compared to mice who did not receive MSCs having engraftment of 25% (p=0.005). Transplantation of double UCBT did not improve engraftment when compared with a single UCBT of equivalent dose. Additionally, the quality of engraftment was enhanced with MSC co-transplantation producing superior engraftment of CD34+ cells. In conclusion, at equivalent cell dose single and double UCBT lead to similar engraftment, suggesting the enhanced engraftment seen with double UCBT reflects a cell dose effect. MSC co-transplantation enhances engraftment of both single and double unit cords and may be a potential strategy to be explored in the clinic.

Infusion of allogeneic mesenchymal stromal cells can delay but not prevent GvHD after murine transplantation M. Kambouris *, B. Turner, L. Sinfield, H. Cullup, D. Hart, K. Atkinson, A. Rice Mater Medical Research Institute (South Brisbane, AU) Multipotent, mesenchymal stromal cells (MSC) are emerging as a means of immunosuppression for patients with steroid refractory graft-versus-host disease (GVHD). Despite clinical use, pre-clinical data is still lacking. We established an in vivo model using MSC to control GVHD to determine their mode of immunosuppression. We showed in a mixed lymphocyte reaction that MSC are highly immunosuppressive and significantly reduce T cell proliferation. We then examined the effects of donor-derived intraperitoneally (IP) or intravenously (IV) injected MSC on GVHD in a myeloablative conditioned, full mismatched model of haematopoietic stem cell transplantation (HSCT) [UBI-GFP/Bl6(H-2b)->BALB/c (H-2d)]. 4x10 5 donor-derived MSC/mouse were injected 4hrs pre or 24hrs post HSCT then mice were monitored daily for GVHD. Only mice given MSC IP 24hrs post HSCT showed a signficant delay in death from GVHD, where median survival was increased by 10 days (day 7 vs day 17, p<0.001, (fig 1) . We then investigated if MSC delivered IP pre or post HSCT altered cytokine-driven trafficking of cellular effectors known to play a role in GVHD via timed sacrifice. Control mice (HSCT and no MSC or MSC and no HSCT) were also sacrificed daily for 6 days following HSCT. We found that MSC given post-HSCT enter an environment of significantly increased activated dendritic cells (DC) in the spleen ( fig 2) and at day 3 and day 6 post HSCT, mice given MSC pre HSCT had increased levels of activated splenic DC compared to mice treated with MSC 24hrs later. At day 3, we also saw more IFN-gamma in spleen washings in mice treated pre HSCT compared to controls. We then determined the role of MSC in GVHD control in a minor mismatch model [UBI-GFP/Bl6 (H-2b)->BALB.B (H-2b)]. MSC were given on day 1, 7 or 14; or to mice with established GVHD (>25% weight loss with other surrogate GVHD markers more than 14 days post HSCT). Mice given MSC therapy after GVHD onset showed no increase in survival compared to controls. However, MSC administered prophylactically at day 1 (4x10 5 /mouse) or day 7 (1x10 6 /mouse) but not at day 14, showed significantly increased survival compared to controls suggesting that the timing of MSC administration is important for their impact on GVHD. In summary, IP injection of MSC influences GVHD and survival and may have in vivo influence on activated DC. Elucidation of the mechanism by which MSC control GVHD may ultimately lead to wider application of their use to assist HSCT. S15 O148 Immune reconstitution after cord blood transplantation by direct intrabone injection of cells A.M. Raiola*, A. Ibatici, V. Pinto, A. Kunkl, F. Guialn, F. Gualandi, D. Occhini, A. Dominietto, C. Di Grazia, S. Bregante, T. Lamparelli, M. Mikulska , G. Piaggio, M. Podestà, F. Frassoni, M. van Lint, A. Bacigalupo S. Martino's Hospital (Genoa, IT) Introduction: cord blood transplantation (CBT) has been increasingly used to treat hematological malignancies. Since March 2006 we have been investigating a pilot study of CBT in adults by injecting CB cells directly into the bone to overcome the cell dose barrier. Pre-clinical transplant models have shown that intrabone (IBM) injection reduces the incidence of GVHD in MHC disparate donor/recipient pairs, suggesting that the way of transplant may affect the immune reconstitution. We have analysed immune reconstitution (T/B/NK cells) at different time-point after IBM-CBT in adults recipients. Materials & Methods: Thirty-three consecutive patients with advanced haematological malignancies a single-unit graft CBT. Median age was 35 years (18-62) and follow-up was 8 months (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) for 27 patients survivors longer than 30 days after CBT. Median cell dose was 2.6 TNC/kg (range, 1.5-4). Donor/recipient HLA matching was highly disparate: 5/6 in 27%, 4/6 in 66%, 3/6 in 7%. All but 5 pts received a myeloablative conditioning regimen (TBI 1200cGy in 26/28 pts). GVHD prophylaxis consisted of CSA, MMF and pretransplant ATG. Immunophenotype on peripheral-blood lymphocyte subsets was assessed on day +30, +60 +100, 6 months and 1 yr after CBT. Results: Early death occurred in 7 pts. All 26 were evaluable and engrafted 100% donor. Overall survival was 50% and relapse-related death was 15% (5 pts). Median time to neutrophil and platelet recovery was 23 and 40 days, respectively. Grade II acute GVHD was seen in 7 cases (27%) and grade III in 1 case (4%). Moderate/extensive chronic GVHD developed in 6/26 (23%). On day +30, median CD4+ and CD8+ count were as low as 11 and 7 cells/mmc, respectively and remained below normal values up to 6 months. At 1 year, CD4+ and CD8+ count normalized (397 and 250 cells/mmc). NK values reached values as high as 381 and 465 cells/mmc on day +30 and +60, then remained stable up to 1 yr. B cells were extremely low up to day+60, peaked to 445 on day +100 and slighty expanded over time. Conclusion: Data from our series showed that immune reconstitution is as delayed as in patient who received CB grafts intravenously, with prolonged T-cell lymphopenia and Bcells spike since day+100 (Komanduri, Blood 2007). We also confirmed early NK expansion which could be due to the high donor/recipient pair HLA disparity associated with the low CD4+ count. If this may have an impact on the relatively low relapse incidence observed in advanced patients warrants further studies.

The effect of mesenchymal stem cells in a rat model of experimental arthritis is dose-and time-dependent E. Yannaki (1) , A. Papadopoulou* (1), M. Yiangou (2) , E. Athanasiou (1) , I. Batsis (1) , A. Athanasiadou (1) , A. Xagorari (1) , A. Anagnostopoulos (1) , A. Fassas (1) (1)George Papanicolaou Hospital (Thessaloniki, GR); (2)Aristotle University (Thessaloniki, GR) Mesenchymal stem cells (MSCs) have recently generated a great deal of excitement, due to their potential of differentiation into a broad spectrum of tissues and immune regulation by non-MHC-restriction. The latter capacity has resulted in an effective control of GvHD in several studies; however, the role of MSCs in autoimmune diseases has not been extensively investigated in animal models. We aimed to explore the effect of rat MSCs on cultured fibrobIast-like synoviocytes (FLS) and T-cells from the spleen after adjuvant arthritis induction (AA) as well as their in vivo immunomodulatory potential in a rat model of AA resembling human rheumatoid arthritis. Culture of AA-FLS in the presence of supernatant from syngeneic MSC culture when compared to FLS in the absence of supernatant, reduced the AA-FLS proliferation (p=0.00004) as well as the proliferation of ConA-stimulated AA T-cells (p=0.04). Coculture of activated T-cells with syngeneic MSCs produced stronger inhibition (p=0.004) in a dose-dependent manner. The inhibitory effect of allogeneic MSCs in activated AA T-cells was even stronger and similarly dose-dependent, either by secreted agents (p=0.017) or by cell to cell contact (p=0.00017). Low doses of MSCs (0.5-5x10 5 cell/recipient) administered iv, intrasplenic or intrabone marrow, at single or multiple infusions, didn't produce significant differences in disease scores of MSCtreated as compared to control rats. On the other hand, repeated, higher dose (6x10 6 cell/rat), iv infusions of syngeneic or allogeneic Y+MSCs to female recipients, before the onset of AA (d4 and d9 post AA), resulted in significantly lower arthritic scores with maintenance of a rather normal joint architecture with mild focal synovial hyperplasia/pannus formation and reduced abnormal chondropIasia or bone erosion as compared to control rats. In contrast, symptom worsening occurred when the same cell dose was injected after arthritis onset (d13 and d20 post AA). At the time of sacrifice (d30 post AA), no Y+MSCs were detected in the spleen or in cultured FLS from synovial membrane, by PCR or FISH, suggesting that the observed benefit was not due to MSCs homing to the target tissues or that migration of MSCs may have happened earlier. Our data suggest MSCs as a potential new therapeutic approach for autoimmune arthritis. However, the benefit of infused MSCs seems to be dose-and time-dependent and further studies are required before the results can be clinically translated. Objectives: Mesenchymal stem cells (MSC) suppress alloantigen-induced proliferation, interferon-gamma (IFNgamma) production and cytolytic killing in vitro and infusion of third-party MSC appears a promising therapy for acute GVHD. However, little is known about the specificity of immunosuppression by MSC and in particular the effect on cell-mediated immunity to infectious pathogens. We have studied the effect of MSC on virus-specific T-cell responses. Results: Peripheral blood mononuclear cells (PBMC) from 6 normal donors were stimulated for 5 days with autologous lymphoblastoid cell lines (LCL) (EBV), pp65 peptides (CMV), an adenoviral vector Ad5f35 (Ad) or allogeneic PBMC (Allo), in the presence or absence of third-party MSC (MSC/PBMC ratio 1:10). MSC significantly suppressed proliferation in response to Allo (mean 61% suppression, p=0.003), but had less effect on the response to EBV (mean 42% suppression, p=0.016) and no suppression of the response to CMV or Ad. MSC had no effect on expansion of EBV and CMV pentamerspecific T cells after stimulation with their cognate antigen. ELISPOT assays demonstrated that MSC significantly inhibited IFN-gamma production in response to Allo (mean SFC/10 5 cells 1125 ±274 without MSC, 263±49 with MSC) and to a smaller extent to EBV (3181±548 vs 2147±387), but not to CMV (2535±374 vs 2532±311). Established EBVspecific cytotoxic T-cells (EBV-CTL) from 5 donors were stimulated with autologous LCL with or without MSC for 5 days. The percentage of EBV-pentamer specific CTL was not affected by the presence of MSC (mean 1.59% vs 1.52%). MSC did not inhibit IFN-gamma production by EBV-CTL (6039±1644 vs 5885±1608). Furthermore, cytolytic killing of LCL by EBV-CTL was not suppressed by MSC (mean specific lysis 47±3.6% vs 52±4.7% at E:T ratio 30:1). Finally, we studied anti-CMV immunity in 2 patiens who received MSC for acute GVHD of the gut evolving into chronic GVHD, with a good transient response. PBMC from these patients showed persistence of pp65-pentamer positive T-cells and retained IFN-gamma response to CMV post MSC infusion (SFC 69/10 5 PBMC pre-MSC, 297 at 1 month and 231 at 3 months). Conclusion: MSC have little effect on T-cell responses to CMV, EBV and Ad, which contrasts to their strong immunosuppressive effects on alloreactive T cells. These data have major implications for immunotherapy of GVHD with MSC and suggest that the effector functions of virus-specific T-cells may be retained after MSC infusion.

Intracerebral transplantation of human mesenchymal stem cells after hypoxic-ischaemic brain damage in rat H. Huang* (1), J. Fan (1), X. Lai (1) , Q. Yu (2) , W. Ding (1) , Y. Wang (2) (1)Zhejiang University School of Medicine (Hangzhou, CN); (2)Zhejiang Chinese Medical University (Hangzhou, CN) Mesenchymals stem cells (MSCs) have dramatic potential of proliferation and multiple differentiation, and recently, neural transdifferentiation of MSCs provoke extensive interest. Our study was designed to explore the migration, differentiation and the therapeutic benefit of hMSCs after hypoxic-ischemic brain damage in rats. hMSCs were prelabeled with bromodeoxyuridine (BrdU) for 72h before transplantation. Animal models of hypoxic-ischemic brain damage (HIBD) were built in one month old Wistar rats, and three days after hypoxia-ischemia, HIBD rats in hMSCs-treated group (n=18) received intracerebral transplantation of 5±10 5 hMSCs, while HIBD rats received PBS of the same volume in control group (n=18). In sham-operated group (n=6) and HIBD group (n=6) which just had hypoxic-ischemic brain damage but did not receive any treatment. All of the groups did not receive any immunosuppression agents. Behavior tests (alternative electro-stimulus Y-maze, by single blind method) indicated that pace memory capacity of rats in hMSCs-treated group is significantly better at 4 weeks post-transplantation. HE straining also showed that the damage caused by hypoxicischemic had lager pathological changes with bulks of neural cells necrosis and neuropil cavitations formation, even disappearance of normal architecture in cortex and hippocampus, however, it would be significantly improved in pathology in hMSCs-treated group. hMSCs stayed around the local area of transplantation in three days after transplantation, then migrated mainly along the ventricular system. Four weeks later, hMSCs were observed to migrate to the parenchyma and distribute throughout the cerebra. Three days after transplantation, some hMSCs expressed GFAP in the local area of transplantation, and a few of cells expressed Neurofilament (NF) near blood vessel. Four weeks after transplantation, BrdU and GFAP co-express cells increased which mainly distributed in the local area of transplantation, the cortex, hippocampus and ependymal layer. Nestin positive cells did not be detected at any time point. Conclusion: After being transplanted intracerebrally, hMSCs could migrate to the lesion area, express NF and GFAP which indicated their differentiation into mature neurons and neurogliocytes, and promote tissue repair and functional recovery of HIBD. Furthermore, these results suggested that hMSCs could be a promising treatment for hypoxic-ischemic brain damage.

Identification and enrichment of human bone marrow derived mesenchymal stem cells by monoclonal antibody, ZUC3 X. Lai, H. Huang*, L. Huang, J. Cao, F. Zeng, J. Zheng Zhejiang Uninversity School of Medicine (Hangzhou, CN) Human mesenchymal stem cells (MSCs) could be well isolated and expanded from bone marrow and have been widely studied, however, there is no specifically definitive marker of MSCs till now. In our previous study, a novel murine monoclonal antibody (McAb) ZUC3 was produced by hybridoma technology, which was specifically reactive with human MSCs, while showed negative cross-reactivity when screened against a variety of human tissues. Flow cytometric analysis showed that ZUC3 antigen expression by cultured MSCs and mononuclear cells derived from bone marrow were 91.31±2.92% and 0.96±0.28% respectively, and western blotting demonstrated the molecular mass of antigen was about 33KD. ZUC3 antigen positive and negative cells were separated from bone marrow mononuclear cells by immunomagnetic activated cell sorting, and plated respectively in human MSCs medium consisting of 10% FBS, LG-DMEM. The purity of the recovered fractions for ZUC3 by MACS was 76.82±6.32% by flow cytometry. The positive cells have adhered to culture flask in vitro, and the quantity of adhered cells that had fibroblast-like morphology increased and proliferated during primary expansion period, while the negative cells were observed as round shape cells without any proliferation. It was demonstrated that ZUC3 antigen positive cells continued growth with spindle-shape, expansion in long-term culture. Phenotype of ZUC3 antigen positive cells was analyzed by flow cytometry, the culture-expanded positive cells were uniformly positive for CD29, CD44, CD105, CD106, and lack typical hematopoietic antigens such as CD14, CD34, CD45, HLA-DR, which demonstrated that ZUC3 positive cells sorted from bone marrow mononuclear cells were MSCs. With proper medium, the ZUCs antigen positive cells could be successfully induced to differentiate into adipocytes, osteoblasts, and neuro-like cells which were positive of neuron markers such as nestin, NSE and NF-M. Conclusion: ZUC3 McAb was a specific surface marker against human MSCs for cell sorting. The ZUC3 antigen positive cells separated from bone marrow mononuclear cells had potential capacity of high proliferation and multiple differentiation.

Everolimus enhances the immunomodulatory properties of CD271 positive selected human mesenchymal stem cells R. Racila*, W. Melchinger, J. Finke, R. Marks Medical Hospital University of Freiburg (Freiburg, DE) Immunomodulatory properties of human mesenchymal stem cells (MSC) suggest their use as a potent cellular therapeutic agent for acute graft versus host disease in allogeneic hematopoietic stem cell transplantation (HCT). Everolimus is a mTOR inhibitor which is currently being tested for GvHD prophylaxis, but its effect on MSCs is not yet understood. Since mTOR inhibition by everolimus might interfere with protein biosynthesis, we tested whether MSCs maintain their immunomodulatory properties in the presence of everolimus. In order to obtain a homogenous MSC population, positive selection of CD271+ cells from bone marrow aspirates from volunteers using anti-CD271 antibody coated magnetic beads were used. The one-step procedure resulted in an enrichment of CD271highCD73+CD45lowCD34-MSC with copurification of CD271lowCD73-CD45high hematopoietic cells. Expansion culture resulted in a pure CD271lowCD45lowCD73highCD105highCD34low MSC population. We assessed the immunomodulatory properties of CD271lowCD45lowCD73highCD105highCD34low

MSCs when cocultured with autologues and allogeneic T cells stimulated with anti-CD2/CD3 antibodies or third party peripheral mononuclear cells. The presence of MSC in allogeneic T cell cultures did not result in a proliferative T cell response. Moreover, stimulation of allogeneic MSC/T cell cultures with anti-CD2/CD3 antibodies resulted in suppression of T cell proliferation as determined by thymidine incorporation. A 65-70% reduction of maximal T cell proliferation was seen when a ratio of 1:10 (MSC to T cell) was used. In order to examine the effect of mTOR inhibition onto the immunomodulatory properties of MSC, we incubated the MSC for 4 hours with 10 nM everolimus. After several washes the everolimus treated and irradiated MSC were used for coculture (ratio 1:10) with stimulated allogeneic T cells. Interestingly, everolimus treated MSCs showed enhanced suppressive properties and a 85-90% reduction of maximal T cell proliferation. We conclude that everolimus enhances the immunomodulatory properties of CD271 positive selected human MSCs and the clinical use of this mTOR inhibitor might result in significant immunsupressive activity in allogeneic HCT.

Fibrosis regression after autologous haematopoietic stem cell transplantation in systemic sclerosis L. Vija, F. Verrecchia, O. Verola, A. de Raignac, D. Sibon, L. Michel , D. Farge St Louis Hospital (Paris, FR) Background: Significant regression of clinical skin sclerosis, as assessed by repeated measure of the Rodnan modified skin sclore (mRSS), has been shown after Autologous stem cell transplantation (ABMT) in diffuse scleroderma (SSc) patients. Objectives: To analyse wether 1) the clinical and histological extent of skin fibrosis can regress after HSCT and 2) extra cellular matrix organization and pro-fibrotic signals elicited by TGF-b in SSc fibroblasts derived from skin biopsies vary according to the modified Rodnan skin score (mRSS) before and after HSCT . Methods: 38 SSc patients underwent 75 skin biopsies with simultaneous measure of mRSS before or after treatment by immunosuppressive drugs with or without autologous HSCT. The histological presence and distribution of sclerosis was assessed as : 0 = no fibrosis, + = light fibrosis, ++ = moderate fibrosis, +++ = extensive fibrosis within the papillary (PD), the superficial reticular dermis (SRD), the median reticular dermis (MRD) and the deep reticular dermis (DRD). Human fibroblasts were established by explanting skin punch biopsy (SSc patients or healthy controls) in Dulbecco's modified Eagle's medium and cells were used for experiments between passages 3 and 8. Immunoblotting analyses with anti-phospho Smad3 (Mayo Clinic, Rochester, MN), anti-Smad3 (Clinisciences, CA), and anti-bactin antibodies (Santa Cruz Biotech, CA) and rt PCR for COL1A1, COL1A2 and PAI-1 have been performed. Results: Double blind optic microscopy analysis of the biopsies seriate sections on standard matrix stains allowed to define 3 histological subgroups : 9 with grade 1 weak fibrosis, 24 with grade 2 moderate fibrosis and 42 with grade 3 severe fibrosis with significant correlation (p< 0.0001) between the grades of fibrosis and the mRSS. In skin fibroblast cultures, Smad3 phosphorylation levels, representative of TGF-b receptor activity, increased in parallel with the mRSS. When compared to pretransplant values, a significant regression of the degree of fibrosis was observed both in the papillary and in the reticular dermis after HSCT (n=7), correlated with a decreased mRSS. Conclusion: In SSc patients, we confirm that the histological extent of skin fibrosis correlates closely with the mRSS and further demonstrate that both parameters regress after treatment by HSCT. The extent of TGF-b signaling activation in SSc skin fibroblasts appears to parallel the severity of disease. We report here an update of the phase 2 multicenter, prospective study of high-dose immunosuppressive therapy and autologous haematopoietic cell transplantation in 21 nonprimary progressive MS patients showing high disease activity on the basis of both brain magnetic resonance imaging (MRI) and sustained clinical deterioration despite conventional therapies. The treatment consisted of stem cell mobilization with cyclophosphamide (4 g/m²) and filgrastim, and conditioning with BCNU (1,3-bis(2-chloroethyl)-1-nitrosourea), cytosine arabinoside, etoposide, and melphalan (BEAM) followed by antithymocyte globulin (ATG, 10 mg/Kg) and infusion of unmanipulated peripheral blood stem cells (PBSCs). The median follow-up is now 66 months (range 21-105). Confirmed progression-free survival was evaluated according to EDSS disability score as the probability of being alive without clinical progression as compared to baseline.

Progression-free survival is 58,16% at 8,5 years after HSCT (figure 1). Seven patients achieved a sustained improvement in their EDSS score, while 6 patients remained stable and 8 worsened of at least 0,5 points in EDSS score. Disease progression was not associated to clinical relapse. Only one patient showed a clinical relapse at +52 months, spontaneously remitted without deterioration of the EDSS score.Since transplantation patients were treated only with symptomatic therapy. HSCT was capable to induce sustained progression and treatment-free survival in a large cohort of rapidly deteriorating MS patient with a modest transplantrelated toxicity. Experimental animal data and single case reports have documented that allogeneic haematopoetic stem cell transplantation (HSCT) potentially alters the course of severe autoimmune disease (AID). Experience with allogeneic HSCT for these indications is still limited. We therefore undertook a retrospective analysis of the EBMT database to identify patients, which received allogeneic HSCT for AID. We identified 31 patients (pts), who underwent 34 allogeneic HSCT for AID between 1984 and 2007. Follow up was requested by a questionnaire, sent to the referring physicians, asking for diagnosis, transplant procedure, complications, disease status and last follow up. The completed questionnaire was available at the time of abstract submission for 15 pts. Median age at transplantation for all pts was 20.07 (0.36-59.3) years; (16 male/ 14 female/ 1 not known). Sixteen HSCT were performed for non-haematological AID and 18 for haematological AID. Three pts had two HSCT for AID. Eight pts had a previous autologous HSCT. Mean follow up time was 37.1 months (IQR 6. 2-85.2) . Responses were complete for 12 pts (35%), partial for 7 (20%), no change for 6 (17%) and unknown for 9 pts (26%). Complete or partial response rate was independent of underlying disease; (haematological vs. non-haematological AID) (p=0.5), gender (p=1.0) or age (p=0.69). Cyclophosphamide based conditioning was associated with a better response rate (p=0,066), while ATG had no influence on the response. Overall mortality was 35%.

Three patients (9%) died of progression of AID, eight patients (25%) died of treatment related reasons. The median followup time of non-survivors (11 pts) was 5.2 months (range 2. 1-28.4 ) and of survivors (20 pts) 67.8 months (range 28. 2-86.1) . Kaplan Maier curves displayed similar survival rates for haematological and non-haematological AID. Best survival was observed for twin donors (3 pts all survived) and matched donors (related or unrelated). TBI conditioning showed a trend to a higher mortality (p=0.09). In conclusion allogeneic HSCT has the potential to induce complete remission in refractory AID. The high mortality rate limits its general use. For selected patients allogeneic HSCT based on cyclophosphamide conditioning from a matched donor could be an option and should therefore be prospectively evaluated. 37 patients with intractable form of MS were included in the phase II clinical trial involving the high dose chemotherapy with autologous peripheral blood progenitor cell (PBPC) support between 1998 and 2006. 33 patients underwent high dose conditioning BEAM. T cell depletion in vitro was performed on 20 grafts depending on number of harvested progenitors and available resources. 13 patients with not purged graft received ATG 4mg/kg i.v. D+1, D+2 after PBPCT. In 3 patients PBPC mobilization failed (cyclophosphamide 4g/m 2 + G-CSF). One patient refused transplantation after improvement in disability following mobilization. Median follow-up is 60 months (12 -108). Median EDSS (Expanded Disability Status Scale) of grafted patients at the time of inclusion was 6.5 (5.0-8.5), median EDSS of grafted patients at last follow up was 7,0 (6.0-10.0). Two patients died 31 and 58 months resp. after transplantation because of progression of MS and the cause not related to transplantation, respectively. Two patients were lost for follow up. There was no treatment related mortality. At last follow up, the significant improvement (by 1.0 point and more on EDSS) remains in 1 patient, stabilisation of the disease occured in 23 patients (70%), 9 patients gained disability significantly (by 1.0 point and more on EDSS). In 3 patients occured transient significant neurological improvement lasting 12-30 months. The development of disability between the group grafted with in vitro purged graft and the group with ATG i.v. was not significant.

Cumulative survival without significant deterioration was 80% in the group with purged grafts and 64,3% in the group without purging in vitro (p=0,178). Among 4 mobilized but not transplanted patients 1 improved by 1.5 point on EDSS, 3 patients worsened by 1.0 point. Three serious adverse events related to transplantation were observed: respiratory failure after the onset of mucositis in the patient with severe pontomesncephalic impairment, sepsis with respiratory failure following bilateral pulmonary hemorrhage, both patients needed temporary artificial ventilation and recovered. Bleeding related to the inhibitor of FVIII occured in 1 patient, remmission has been achieved after the immunosuppressive therapy. As majority of patients with otherwise intractable MS at least stabilized in their disability, we consider the results to be promising and requiring confirmation in a randomized trial involving also less handicapped patients.

Long-term results of a GITMO retrospective study on haematopoietic stem cell transplantation for paroxysmal nocturnal haemoglobinuria S. Santarone*, E. Di Bartolomeo, A. Bacigalupo, E. Tagliaferri, A. Iori, A. Risitano, S. Tamiazzo, F. Papineschi, A. Rambaldi, A. Spagnoli, E. Angelucci, P. Di Bartolomeo on behalf of the GITMO Allogeneic haematopoietic stem cell transplantation (HSCT) may cure paroxysmal nocturnal hemoglobinuria (PNH). In this study we report the results of allogeneic HSCT in 26 patients (16 males and 10 females) affected by PNH who were transplanted between July 1988 and May 2007. The median age at time of HSCT was 32 years (22-60). The median time from diagnosis to HSCT was 33 months (3-208). All patients had received various treatments before HSCT including steroids, immunosuppressive drugs and growth factors. Twenty-one patients were transfusion-dependent. The median number of packed red blood cells and platelet concentrates received before HSCT was 30 (4-500) and 33 (6-86) respectively. The median peripheral hematological counts at transplant were: polymorphonucleates (PMN) 1780 (20-10400) x10 9 /L, hemoglobin 8,7 g/dl (4.6-11), platelets (PLT) 79

(6-355) x10 9 /L. Thirteen patients developed thromboembolic episodes before HSCT. Twenty-four patients were transplanted from HLA identical sibling and 2 from matched unrelated donors. The donor's median age was 33 years (20-59). The conditioning regimen was myeloablative for 15 patients (Busulfan and Cyclophosphamide), whereas 11 patients received a reduced intensity conditioning including Fludarabine, Cyclophosphamide, Melphalan and Total Body Irradiation. As graft-versus-host disease (GvHD) prophylaxis, 11 patients received Cyclosporine (CSA) alone and 13 were given CSA and short course Methotrexate. Two patients received T-cell depleted marrow cells. Twenty patients were given bone marrow cells (median nucleated cells 4.0 (2,2-7,5) x10 8 /kg) and 6 received peripheral blood stem cells (median CD34+ cells 3.5 (1.7-8.4) x10 6 /Kg). Twenty-five patients engrafted with a median time of 17 (10-38) days to reach >0.5 x10 9 /L PMN and 26 days to reach >50 x10 9 /L PLT. The probability of developing grade II-IV acute GvHD and extensive chronic GvHD was 42% and 16% respectively. The transplant related mortality at 6 months was 34%. Causes of death were infection in 4 patients, acute GvHD in 1, chronic GvHD in 2, multi-organ failure in 1 and EBV-related disease in 1. As of December 2007, 16 patients are alive with complete hematological recovery and no evidence of PNH at a median follow-up of 109 months (8-212). The 10-year Kaplan-Meier probability of disease-free survival for the 24 patients transplanted from related donor is 70%. No patient developed thromboembolic disease following HSCT. This study confirms that HSCT is a curative treatment for the majority of patients with PNH.

Subcutaneous alemtuzumab is safe and effective for treatment of global or single-lineage immune-mediated marrow failure: a pilot study from the Working Party Aplastic Anaemia (WPSAA) A. Risitano*, L. Marando, C. Selleri, B. Serio, E. Seneca, A. Camera, L. Catalano, G. Scalia, L. Del Vecchio, A. Iori, S. Maury, A. Bacigalupo, G. Sociè, A. Tichelli, J. Marsh, H. Schrezenmeier, J. Passweg, B . Rotoli on behalf of the WPSAA Acquired marrow failure syndromes may globally affect all hematopoietic lineages, as in aplastic anemia (AA), or may selectively involve single lineages, as in pure red cell aplasia (PRCA) or in agranulocytosis (AGR). Standard treatment for these conditions includes immunosuppression (IS), because of their common cellular immune-mediated pathophysiology.

Here we report a phase II/III pilot study with alemtuzumab (MabCampath®) (ALE) followed by low-dose cyclosporine A (CsA) as alternative IS regimen in 24 patients suffering from marrow failure (13 AA, 7 PRCA and 4 AGR). All patients received subcutaneous ALE as escalating dose in consecutive days (3-10-30-30-[30] mg, total dose 103 for AA and 73 for PRCA and AGR), premedicated by betamethasone, clorpheniramine and paracetamol. Six patients received one or more additional courses as a result of relapse, so a total of 32 courses were administered. All patients started oral low dose CsA (1 mg/kg) on day 7. An intensive anti-infectious prophylaxis has been exploited, which included oral valgancyclovir and cotrimoxazol as anti-CMV and anti-P. Carinii agents, respectively. All patients completed the treatment with no relevant injection-related side effect (with exception of fever in some cases), nor significant clinical or laboratory abnormality. A complete lympho-ablation was observed in all patients within 2-3 days, which persisted for several weeks; transient worsening of neutropenia and/or thrombocytopenia were observed in some patients. At a median follow-up of 9 months, infectious events were irrelevant: in cumulative 184 patient-months, 2 HSV and 1 flu have been recorded, all resolving quickly. No CMV reactivation was demonstrated. Immune reconstitution was delayed up to several months, especially affecting the CD4+ compartment. Patients with adequate follow up (more than 3 months) were assessed for treatment efficacy: 8 AA patients showed 2 CR, 4 PR, 2 NR. In the 5 PRCAs, there were 4 CR and 1 NR; 3 out of 3 AGRs obtained CR. Among responding patients, 2/3 SAAs, 3/4 PRCAs and 1/3 AGR experienced relapses, which were successfully treated by additional course of ALE. In conclusion, subcutaneous ALE is a feasible and safe IS regimen for patients suffering from immune-mediated marrow failure syndromes. Preliminary results suggest excellent response rate, and support efficacy even in case of relapse; such favorable risk-to-benefit ratio predicts for this regimen a leading position in the future IS strategies. Stable mixed chimerism has been described as being beneficial in patients with aplastic anemia after stem cell transplantation, as these patients were protected from severe graft versus host disease and from rejection, the major drivers of mortality in these patients. Patients with marrow failure syndrome do not need full donor chimerism as they do not benefit from a graft-versus-malignancy effect. Here we describe a protocol resulting in high rate of stable mixed S20 chimerism in patients conditioned with cyclophosphamide + ATG and transplanted with peripheral stem cells depleted by Campath with add-back of T-lymphocytes post-transplant at different dose levels, according to donor-recipient relationship. We included 10 patients in this protocol, median age was 35,5 (15-58) years, 5 were male, donors were identical siblings in 7, mismatched related in 1 matched unrelated in 1 and mismatched unrelated in 1, respectively. Stem cell dose was 11.5 (5.4-40.2 ) CD34 x 10 6 /kg. Time to neutrophil engraftment was 15 (8-17) days and to platelet engraftment 14 (8-24) days. Median follow-up of surviving patients is 627 days (13-2533) Acute grade II GvHD was observed in 1 patient (with the mismatched unrelated donor), this patient died subsequently, while all other patients remain alive. There is mild chronic GvHD in 1, none of the patients lost the graft. Chimerism analysis showed full donor chimerism in 2, transient mixed chimerism (defined as mixed chimerism developing into full donor chimerism) in 1, and stable mixed chimerism in 6, limited to mononuclear cells in 1 and observed in the granulocyte and mononuclear compartment in 5. Mixed chimerism remained stable for up to 4 years. A protocol of conditioning with cyclophosphamide and ATG and GvHD prophylaxis with T-cell depletion using Campath and T-cell addback may induce stable mixed chimerism in a large proportion of patients with aplastic anemia with low GvHD and graft failure risks.

I. Resnick* (1), A. Maschan (2), M. Shapira (1), P. Trakhman (2), E. Skorobogatova (2), D. Balashov (2), M. Aker (1), R. Or (1) (1

Introduction: Ten years ago we introduced a fludarabine (Flu) based conditioning for hematopoietic stem cell (HSC) and umbilical cord blood transplantation [1, 2] for Fanconi anemia (FA). This approach dramatically improved results of FA treatment. Following initial case reports and small series [3, 4] , two large studies were published: by an Italian group and of European experience [5, 6] . Aim: Here we present an update of our two centers' experience of HSC transplantation for FA. Methods: Thirty patients (age 3.2 to 31, median 10.3 years) underwent allogeneic HSC for treatment of FA. Used conditioning protocols were Flu based in 19: Flu 150 or 180 mg/m 2 + cyclophosphamide (Cy) 20 mg/kg or busulfan (Bu) 4 mg/kg or both + antithymocyte globulin (ATG) Atgam 90 mg/kg or Fresenius 40mg/kg. An alternative conditioning (11 patients) was Cy-TAI/TLI or Bu8Cy40. Follow up till December 1, 2007 was 3 to 112 months. Results: In the group of patients who got Flu based protocols disease free survival (DFS) is 84% vs 27% in pre-fludarabine protocols; p=0.006; OR=14.2 (2.5-82.1). Difference of DFS in patients transplanted from matched family donor (92%, n=12) vs matched unrelated donor (71%, n=7) was insignificant. Causes of death were acute graft-vs-host disease (GvHD) grade ≥2 (n=6) in combination with sepsis (n=3), multiorgan failure (n=4), liver veno-occlusive disease (n=3), bleeding (n=2), chronic extensive GvHD (n=2), secondary malignancy (n=1). There was no clear advantage in using a combination of Cy and Bu compared with single agent protocols (DFS 3 out of 4 patients). Three patients suffered a rejection and underwent a 2nd transplant (2 are alive and well). All surviving patients are disease free and in an excellent clinical condition. Conclusion: Combination of fludarabine with ATG and low dose Cy and/or Bu is safe, demonstrates low rejection rates and is well tolerated by FA patients. Lower doses of Flu (150 mg/kg) can be safely used without a pronounced risk of rejection. References: 1.Kapelushnik J et al. Bone Marrow Transplant. 1997; 20:1109 . 2.Aker M et al. J Pediatr Hematol Oncol. 1999 21:237. 3.Bitan M et al. Biol Blood Marrow Transplant. 2006; 12:712. 4.Maschan AA et al. Bone Marrow Transplant. 2004; 34:305. 5.Gluckman E et al. Biol Blood Marrow Transplant. 2007; 13:1073 . 6.Locatelli F et al. Haematologica 2007 92 (10) Results: CB recipients were more likely to have advanced leukemia at conditioning (relapse or induction failure, CB vs. BM = 47% vs. 31% in AML patients, and CB vs. BM = 28% vs. 21% in ALL patients, p<0.0001, p=0.087, respectively). The proportion of ALL with Philadelphia chromosome abnormality was higher (38% to 25 %, p=0.001) among CB recipients. Human leucocyte antigen (HLA) was serologically mismatched in 93% of patients with AML and 93% of patients with ALL among CB recipients, while HLA A, B, and DR were all matched genotipically for BM recipients. In controlled comparisons using multivariate analyses, among patients with AML, higher rate of treatment-related mortality (TRM) (hazard ratio [HR]=1.51, 95% confidence interval [CI], 1.11-2.05, p=0.008) was observed in recipients of CB, which contributed to decreased overall survival (HR=1.45, 95%CI, 1.14-1.84, p=0.003). Relapse rate did not differ between the two groups of AML patients (HR=1.27, 95%CI, 0.91-1.79, p=0.16). In patients with ALL, the relapse rate was higher with marginal significance among CB recipients (HR=1.45, 95%CI, 0.98-2.14, p=0.064). There was no significant difference between these groups for TRM (HR=1.36, 95%CI, 0.94-1.95, p=0.10) or overall mortality (HR=1.25, 95%CI, 0.94-1.67, p=0.12) for patients with ALL. There was no increase in the incidence of acute graft versus host disease in CB recipients among patients with either AML or ALL despite HLA disparity. Conclusions: Matched or mismatched CB is a favorable alternative stem cell source for patients without a suitable donor. We found different outcomes between patients with AML and ALL, which indicates the importance of disease-specific analyses in alternative donor studies. Decreasing TRM is required to improve the outcome for CB recipients, particularly for patients with AML. . The original Minneapolis conditioning regimen was used in 106 (96%) pts and modified in 6 (4 or 6 Gy TBI: 4 pts; ATG : 2). Characteristics of the grafts: A single unit was infused in 77 pts (69%), two in 35 (31%). HLA compatibility was 6/6 in 6 pts, 5/6 in 36, 4/6 in 60, ≤ 3/6 in 6; 43 pts were ABO matched. Infused nucleated cells (NC) was 3.1x107/kg (1-9): 2.9 x 107/kg in single units and 3.7 x 107/kg in double units. CsA and MMF were used for GVHD prophylaxis in all pts. Results: Neutrophils recovery was 85±4% at a median of 19 days (0-48) ; 14% pts experienced autologous recovery; 14% had mixed and 72% full donor chimerism at D+100. Univariate analysis indicated the low weight, previous transplantation, double units and HLA compatibility as significant factors for neutrophil recovery; however multivariate analysis did not find any significant factor. Acute GVHD was observed in 34±5% of pts: 21, 12 and 5 pts had grade II, III or IV aGVHD respectively and chronic GVHD in 16%. Non relapse mortality was 12±3% at 6 months; relapse: 22±5%; overall survival: 72±5%. Causes of death were relapse in 17 pts, GVHD in 2 pts, venocclusive disease and multiorgan failure in 5, infections in 4 and other toxicity in 3. DFS at 6 and 24 months were 68±5% and 65±5%, respectively. Multivariate analysis identified 3 independent risk factors: HLA disparity(0+1 vs 2+3), cell dose (<3.1x 107/kg) and age (>44y). This assessment of UCBT after NMA confirms the good results of the Minneapolis group (Brunstein et al. Blood 2007) . Few events were observed between 6 and 24 mo and DFS remains high and UCBT is a good option in absence of other source of stem cells.

Background: Cord blood transplants (CBT) are associated with delayed or failed engraftment in a significant proportion of patients. We hypothesized that direct intrabone (i.b.) transplant of cord blood cells could improve hematologic recovery. Methods: Unrelated CB cells were selected for 37 consecutive patients (25 patients. 4/6, 11 patients 5/6 HLA and one pt. 3/6 antigen matched). Median transplant cell dose was 2.5 x10 7 /kg (range 1.4 -4.2). CB cells were concentrated in 4 syringes of 5 ml each and infused in the supero-posterior iliac crest (SPIC) (11 patients bilaterally; 26 patients monolaterllay) under rapid general anesthesia (10 min. with propofol). Patients' median age was 40 years (18-63); 30 had acute leukaemia (21 with refractory or relapsed disease and 6 high risk first remission leukemia); 2 chronic myeloid leukemia in advanced phase; 2 refractory Hodgkin's disease; 2 LNH in advanced phase and 1 aplastic anemia/MDS. Most patients (n=28) were prepared with conventional TBIcyclophosphamide. Results: The infusion of cells i.b. in SPIC was uneventful. Six patients are not evaluable because they died of multiorgan failure within 14 days from transplant. The patient with SAA/MDS did not engrafted and was re-transplanted. All the other patients surviving more than 14 days engrafted (100%). Median for PMN engraftment (>0.5x10 9 /l) was day 25 (14-44), whereas for platelets (>20x10 9 /l) it was day 40 (range 22-64). Four patients died of infection; one patients died of PTLD on day +140. Four patients relapsed and died. Twenty-two patients are alive in remission at a median follow up of 9 months (range 2-21). From day +30 full donor chimerism was documented in CD3, bone marrow cells and progenitor cells from both the injected and non-injected SPIC; from day +30, CFC progenitors reached the values of normal individuals in bilateral sites documenting the colonization of the hematopoietc system and possibly an improvement of seeding efficiency. Only 4 patients experienced acute GvHD grade II and 2 grade I; 4 patients have moderate chronic GVHD and one patient extensive cGVHD. Conclusion: Direct intra-bone transplant of CB cells overcomes the problem of graft failure even when low numbers of HLA mismatched CB cells are transplanted. Low incidence/severity of acute GVHD was observed. Nearly all patients for whom a CB unit was searched were able to undergo CBT. This approach may change the policy of hemopoietic cell transplants. Background: The impact of donor-recipient ABO matching on outcomes after allogeneic stem cell transplantation (SCT) has been a matter of controversy. Objective: To evaluate whether ABO matching has a significant influence on overall survival(OS) in patients(pts) receiving SCT for hematologic malignancies. Methods: We conducted an individual patient data-based meta-analysis using a pooled dataset extracted from 7 centers(ctrs) not included in previous pooled analyses. We analyzed pts who had received bone marrow or peripheral blood transplantation for hematological malignancies. The primary endpoint was OS, comparing pts receiving an ABO matched (MA) graft with those receiving a major (MJ), minor (MI), or bidirectional ABO mismatched (BI) graft using a multivariate (MV) Cox model. In addition, OS and mortality within 100 days were analyzed with adjustment for confounders. Results: A total of 1208 SCT, including 697 MA, 202 MJ, 228 MI, and 81 BI cases, were analyzed. They included 709 related SCT (RSCT) and 184 unrelated SCT (URSCT) from Western ctrs; 214 RSCT and 101 URSCT from Asian ctrs. The median age of recipients was 39 years (range,1-69). The probabilities of OS [95% confidence interval(CI)] at 5 years among MA, MJ, MI, and BI groups were 48% (44-52), 48% (40-56), 45% (38-51), and 37% (26-49) with a median followup of 37 months (range,3-268). Overall, adverse impact on OS was observed for BI group [hazard ratios (HR) adjusted by age and sex:MJ 1.01 (0.80-1.27), MI 1.21 (0.99-1.50), and BI 1.38 (1.01-1.87)]. Among recipients of RSCT, we observed no significant difference in OS between the MA group and any other group. In contrast, MI and BI groups among URSCT recipients experienced worse OS. Sources of heterogeneity were pts who received bone marrow; who had acute leukemia; who underwent transplant at Asian ctrs. In MV analysis of OS and early mortality adjusted for confounders, MI and BI groups showed inferior OS among the subset of URSCT, and an increased risk of early mortality was observed only among MJ group regardless of donor type [HR (95%CI): MJ 1.50 (1.09-2.05), MI 1.23( 0.90-1.70), and BI 1.21( 0.75-1.93)]. Conclusion: Our meta-analysis demonstrates an overall adverse association between BI transplants and survival, largely accounted for by adverse impact of minor or bidirectional ABO mismatching on OS in URSCT. ABO mismatching appeared to have little or no impact on OS in RSCT. Larger studies including URSCT of various ethnic backgrounds should be performed.

Intra-bone marrow injection of umbilical cord blood: no impact on rate of haematopoietic recovery C. Brunstein, J. Barker, D. Weisdorf, T. DeFor, J. Wagner University of Minnesota (Minneapolis, US) Neutrophil engraftment after umbilical cord blood (UCB) transplantation is slower than that observed after adult peripheral blood or marrow transplantation.In order to augment engraftment, we evaluated the safety and potential efficacy of UCB intra-bone marrow injection(IBMI). It was hypothesized that direct IBMI to the marrow microenvironment would reduce hematopoietic stem cells (HSC) loss and improve homing. Based on our prior experience with two partially HLA matched UCB units, the median time to neutrophil engraftment was 23 days (r:15-41). Trial success required: 1) absence of severe adverse events within 48 hours of IBMI, and 2) a 5-day reduction in days to neutrophil recovery (i.e., median ANC>500/mcL by day 18). Based on these measures of success, 29 patients were planned. All received cyclophosphamide 120mg/kg, fludarabine 75mg/m²/day and total body irradiation 1320 cGy with mycophenolate mofetil and cyclosporine A immunoprophylaxis. All pts received 2 UCB units randomly assigned to either IV infusion or IBMI. The IBMI unit was volume reduced to ~20mL with ~10mL infused into each posterior superior iliac crest under local anesthesia at the pts bedside. Ten pts were evaluable (2 were disqualified due to inability to volume reduce the IBMI unit and IBMI unit was not randomized). Median recipient age was 35 years (20-44) and median weight 73.6 kg (56-92). Seven had acute leukemia and 3 had non-Hodgkin's lymphoma. Median infused cell doses for the IV and IBMI units were 2.0x10 7 TNC/kg (1.2-3.0) and 1.8x10 7 TNC/kg (1.4-3.4), respectively. Pts received two 6/6 (n=1), two 5/6 (n=2) or two 4/6 (n=7) HLA matched units. No adverse events were reported with the IBMI procedure. Nine of 10 engrafted. Median time to neutrophil and platelet recovery (>50,000/mcL) was 21 (17-49) and 69(30-272)days, respectively. Complete chimerism was observed with one unit engrafting long term (IBMI unit in 4 and IV unit in 5). Seven of 8 evaluable pts had acute graft-versus-host disease (grade II in 5 and grade III in 2). With a median follow up of 10 months, the probability of survival is 47%(95%CI,14-80%) at 1 year. Based on this interim analysis, the study was prematurely discontinued based on the fact that the predetermined required rate of neutrophil recovery (median 18days) was unlikely with high degree of certainty. While technically easy and safe, neither neutrophil nor platelet recovery after IBMI were not improved to warrant additional patient accrual.

J. Fernandes*, V. Rocha, D. Setubal, M. Bierings, M.A. Champagne, R. Pasquini, G. Socié,

Primary graft failure (PGF) is a fatal complication after allogeneic hematopoietic stem cell transplantation (HSCT). Unrelated donor cord blood transplantation (UCBT) is frequently reported as having a delayed neutrophil recovery and the incidence of PGF varies from 10-25%. Second transplants for PGF have been associated with a high treatment related mortality (TRM) and a poor outcome, with an overall survival (OS) varying from 0 to 30%. We retrospectively analyzed 35 patients who received a second UCBT following PGF after a first UCBT for hematological diseases, between 1995 and 2007. Were defined as PGF patients that did not achieved a neutrophil count >500x10 6 /L until 60 days after UCBT or that received a second transplant for PGF in this period. Patients having relapsed in the first 100 days after UCBT were excluded. Median age was 11.3 years (range, 1-51). Diagnoses were: acute leukemia (n=17), myelodysplasia (n=4) and bone marrow failure syndromes (n=14 -2=Aplastic anemia and 12=Fanconi anemia). Patients received either one (n=31) or two (n=4) unrelated cord blood units matched 3-6/6 (HLA A and B low resolution and DRB1 allelic typing) as donors for the first transplant. Median time between first and second transplants was 55 days (22-116) and median follow-up after the second transplant was 18 months. Concerning the second transplants, 3 patients received mismatched unrelated bone marrow grafts, 7 haploidentical related grafts, 15 single UCB and 10 double UCB grafts. T-cell depletion was used in 8 cases. Conditioning regimen was of reduced intensity in the majority of cases, the most frequent being Fludarabine-based regimens. Low-dose TBI was administered to 10 patients and serotherapy (ALG/ATG/other MoAb) was used in 16 patients. Median time to neutrophil recovery was 19 days (10-47) and 19 of 35 patients engrafted (4 of 15 receiving a single CB unit, 7 of 10 receiving double CB units and 7 of 10 other alternative donors). Chimerism analysis for these patients at engraftment showed: 16=complete donor and 3=mixed chimerism. Twelve patients (34%) developed acute graft-versus-host disease (GVDH) grades II-IV and chronic GVHD was observed in 5 of the 18 evaluable patients (28%). At 1 year, TRM was 39%±1% and OS was 41%±8%. In conclusion, these results suggest that second transplants in this set of extremely high risk patients are effective and should be considered early following primary graft failure after UCBT.

Umbilical cord blood (UCB) is increasingly used as a source of hematopoietic stem cells (HSC) for transplantation in patients (pts) with leukemia. Initial reports suggested that UCB transplantation was associated with a 20-30% leukemia freesurvival (LFS). Therefore, we evaluated the outcomes in 184 consecutive patients transplanted with UCB after a myeloablative therapy between 1995 and 2007 for the treatment of acute myeloid leukemia (AML,n=81), acute lymphoblastic leukemia (ALL,n=84), chronic myeloid leukemia (CML,n=13), and myelodysplastic syndrome (MDS,n=6). Conditioning consisted of either cyclophosphamide (CY) 120mg/kg, fludarabine (FLU) 75mg/m²/day and total body irradiation (TBI) 1320 cGy with mycophenolate mofetil (MMF) and cyclosporine A (CsA) immunoprophylaxis or the same CY/TBI with equine anti-thymocyte globulin (ATG) 90mg/kg and methylprednisolone (MP) immunoprophylaxis. The median age was 16yrs (r: 0.5-52), median weight 57.3kg (r: 8-148). Most were male (57%), cytomegalovirus seropositive (51%). Grafts were composed of two units in 45% and HLA mismatched at one (41%) or two (51%) antigens. The median infused cell doses were 3.4 x 10 7 nucleated cells/kg (r: 0.9-14), 3.9 x 10 5 CD34 cells/kg (r: 0.4-35), 1.2 x 10 7 CD3 cells/kg (r: 0.1-3.2). Pts were classified as standard risk (acute leukemia in CR1-2 or CML in first chronic phase, n=138) or high risk (n=46). Median follow-up for survivors is 2.8 years (r: 0.7-9.2). Engraftment occurred in 164 patients (90% (95%CI, 86-94%). Incidence of grades II-IV and III-IV acute and chronic GVHD was 43%(95%CI,36-50),14% (95%CI, (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) and 15% (95%CI, 10-20%), respectively. Incidence of treatment-related mortality at 2 years and relapse at 4 years was 27% (95%CI, 21-33) and 26% (95%CI, 19-33), respectively. LFS was 43%(95%CI,35-51) and overall survival was 45%(95%CI,37-53) at 4 years. In multivariate analysis, disease risk group was the only risk factor for poor LFS (RR 2.0;95%CI,1.3-3.1;p<.01) and not year of transplant, pt age and weight, NC and CD34 cell dose, treatment regimen, GVHD, CMV serostatus and HLA disparity. Clearly, LFS after UCB transplantation, despite HLA mismatch, represents a valuable alternative to bone marrow and peripheral blood especially for pts requiring urgent transplant or lacking an HLA matched unrelated adult volunteer donor.

Double umbilical cord blood transplantation using reduced-intensity conditioning: a single-centre experience

Double umbilical cord blood transplantation (DUCBT) reduces the time to engraftment and as a result may reduce associated transplant-related mortality. In this report, we describe our experience with DUCBT after reduced intensity conditioning. Methods: 51 subjects were treated on 2 sequential protocols between 2003-2007. All patients received fludarabine (30 mg/m²/d x 6), rabbit ATG (1.5 mg/kg/d x 4) and melphalan (100 mg/m² x 1) as conditioning. GVHD prophylaxis began on day-3 and was administered through day 100 and then tapered. Cyclosporine/mycophenolate mofetil (Cy/MMF) was given to the first 21 patients, and sirolimus/tacrolimus (Siro/Tac) was given to the remaining 30 patients. All patients received 4/6 or better allele level HLA-matched cord units and a supportive care regimen of antiviral, antibacterial and antifungal prophylaxis. G-CSF was administered from day+5 through engraftment. Results: The median age was 49 yrs (19-67) and 30 patients (59%) were male. Malignant diagnoses were similar although there was a higher proportion of advanced lymphoid malignancy in the Siro/Tac group. Subjects received a combined total of 4.4 x 10 6 TNC/kg and 1.9 x 10 5 CD34 cells/kg, without cohort differences. Engraftment kinetics did not differ based on GVHD prophylaxis, with neutrophil engraftment (ANC of 500) at a median of 21 days (13-70), and platelet engraftment (20 000/microL) at a median of 42 days (21-185). 5 patients experienced graft loss between days 35 and 102 (3 Cy/MMF, 2 Siro/Tac). Grade II-IV acute GVHD occurred in 20% (33% Cy/MMF, 10% Siro/Tac, p = 0.05). Chronic GVHD occurred in 24%, without cohort differences. 100-day TRM was 12%, and the risk of relapse at 1 year was 19% (Cy/MMF 5%, Siro/Tac 32%, p = 0.03). Atypical and viral infections were common, and EBV-PTLD occurred in 5 patients. With a median follow-up of 20 months, relapse-free and overall survival at 1 year are 59 and 74%, without cohort differences. Causes of death included infection(6), relapse(5), EBV-PTLD(4), and other(4). Conclusions: DUBCT following reduced intensity conditioning results in early engraftment and low 100-day treatment related mortality. The use of Siro/Tac GVHD prophylaxis is associated with less GVHD than Cy/MMF prophylaxis. This approach may lead to a higher relapse incidence, although this may be explained by differences in subject baseline characteristics. DUBCT should be considered for all adults in whom a suitable adult donor does not exist. (1) RNA encoding the Wilms tumor protein (WT1) is overexpressed in the vast majority of patients with acute myeloid leukemia (AML) and peptide vaccination studies have demonstrated that WT1 is a target for active specific immunotherapy. Preclinical data from our laboratory and that of Hans Stauss have shown that WT1 mRNA-electroporated dendritic cells (DC) stimulate WT1-specific T cells in vitro (Van Driessche A et al. Leukemia 2005; 19:1863 -1871 . Therefore, we started a phase I/II dose escalation trial in which patients with AML received intradermal injections with WT1 RNAloaded DC. Feasibility, safety, immunogenicity and antileukemic activity of the vaccine were investigated. High risk AML patients (all but one of them in complete remission) underwent a leukapheresis and monocytes were isolated using CD14 immuno-magnetic beads by CliniMACS. DC were generated in 6-day cultures in clinical-grade medium supplemented with serum, GM-CSF and IL-4 and matured with PGE2 and TNF-alpha. Keyhole limpet hemocyanin (KLH) was added during maturation as a CD4+ T-helper antigen. Mature DC were harvested, electroporated with WT1 mRNA and used as vaccines. Eight patients received four biweekly DC vaccines. A delayed-type hypersensitivity (DTH) test was performed 2 weeks following the last vaccination. Patients were monitored for minimal residual disease (MRD) by analyzing WT1 RNA expression in peripheral blood by qRT-PCR. Before and after the vaccination cycle, peripheral blood was collected for immunomonitoring purposes. There was successful DC generation and vaccine production in all patients selected. No serious adverse events or toxicity was seen. A decrease in WT1 RNA expression was observed during the course of the vaccination in 4/6 patients who had an increased WT1 mRNA level in peripheral blood at the start of DC vaccination. A vaccine-specific immune response was demonstrated in 8/8 patients by DTH both to KLH as well as to WT1. Preliminary data from immunomonitoring in pre-and post-vaccination T cell samples from 4 patients show a mixed T helper (Th)1/Th2 response towards the KLH and the WT1 protein following vaccination. We conclude that vaccination of AML patients with WT1 RNAloaded DC is feasible and safe. Furthermore, the vaccine elicits anti-vaccine T-cell responses in vivo and a decrease in WT1 RNA expression levels was observed during MRD monitoring in some vaccinated patients, strongly suggestive of antileukemic activity of the DC administered.

Reduced-intensity allogeneic stem cell transplantation using related haploidentical donors for relapsed or refractory lymphomas: evidence of anti-tumour activity in poor prognosis disease P. Corradini*, C. Carniti, A. Raganato, A. Vendramin, L. Farina, V. Montefusco, R. Milani, M. Milanesi, P. Longoni, L. Gandola, C. Lombardo, A. Dodero Istituto Nazionale dei Tumori (Milan, IT) Haploidentical stem cell transplantation (haplo-SCT) has disclosed new possibilities for the treatment of hematologic malignancies in patients (pts) who do not have a HLA compatible sibling or a fully matched unrelated donor. The role of haplo-SCT in relapsed lymphomas is, at present, unknown. We devised a new strategy potentially useful also for pts without NK alloreactive donors (NK alloreactivity in lymphomas is untested). We report the results of a phase I-II trial evaluating early add-backs of CD8-depleted donor lymphocytes (DLIs) (from 1x10 4 up to 5x10 4 cells/kg from day+45 to day +105 at monthly intervals) after a reduced intensity conditioning (RIC) regimen in 22 pts with relapsed lymphomas. Ex-vivo and in-vivo T-cell depletion were carried out by CD34+ cell selection and alemtuzumab, respectively. Histologies were non-Hodgkin's lymphomas (NHL, n=13) [chronic lymphocytic leukemia (CLL), n=5, aggressive NHL (HG-NHL), n=7 and follicular cell lymphoma (FCL), n=1] and Hodgkin's disease (HD, n=9). The median age was 33 years (range, 15-65). Pts received a median number of 4 lines and 86% failed autograft. At median follow-up of 29 months (range, 3-60 months), 10 pts (44%) were alive (n=7 in durable CR) and 12 died of disease (n=8) or non-relapse mortality (n=4). Two-year OS and PFS of the whole group were 44% (95% CI, 22%-64% ) and 33% (95% CI, 12%-56%), respectively. The main factors influencing 2-year OS were age (≤ 45 years: OS of 51%) and chemosensitivity of disease at time of SCT (75% versus 25% for chemosensitive versus chemorefractory, p<0.02). Before DLIs, only 2 of 22 pts (9%) developed acute GVHD. CD8-depletion, perfomed using an immunomagnetic method, reduced the content of CD8+ Tcells by a median value of 3.3 log (1.7-4). A total of 36 CD8depleted DLIs were administered to 16 pts. Three pts (18%) developed acute GVHD grade II-IV following DLIs. Eight pts receiving CD8-depleted DLIs never experienced a relapse. We observed a statistically significant expansion of CD4+ and CD19+, but not of CD8+ and NK cells, at day +120 following CD8-depleted DLIs. Eight pts had measurable TREC/µg at 1 year after SCT. Spectratyping of T cells demonstrated a polyclonal Vß repertoire: pre-transplant 97%, post-transplant 70%. Escalated doses of CD8-depleted DLIs increase the level of CD4+ in the first 4 months after haplo-SCT and are feasible. This is the first report of long-term remissions in lymphomas after RIC haplo-SCT.

K. Perruccio*, F. Topini, A. Tosti, A. Carotti, T. Aloisi, F. Aversa, M. F.

After haploidentical stem cell transplantation, immune recovery is slow due to decaying thymic function and extensive T-cell depletion of the graft which is needed to prevent Graft-versus-Host Disease (GvHD). Consequently, infectious related mortality is about 30-40%. To address this problem, we investigated the efficacy of adoptive immunotherapy after photodynamic purging of alloreactive T cells (Theralux TM , Kiadis Pharma, Amsterdam, The Netherlands) in preventing GvHD and improving immune S25 reconstitution. Optimized protocol conditions provided 3,260 ± 450 (mean ± SD)-fold allodepletion, full retention of Tregulatory cells, and preservation of pathogen-specific T-cell responses (against Aspergillus, Candida, Cytomegalovirus (CMV), Adenovirus (ADV), Herpes Simplex Virus (HSV), Varicella Zoster Virus (VZV), Toxoplasma antigens). Here we present the preliminary results of a clinical trial. Escalating doses of photodynamically allodeleted donor T cells, i.e., 1.25 x 10 5 /Kg, 2.5 x 10 5 /Kg, 5 x 10 5 /Kg and 1 x 10 6 /Kg, were infused into groups of haploidentical transplant recipients. Only 1 patient developed grade III aGvHD at the 1 x 10 6 /Kg cell dose and responded to immune suppressive treatment. Immune assessment analyses revealed that infusion of cell doses equal or greater than 5 x 10 5 /Kg are associated with significant reconstitution of T-cell counts and appearance of pathogen-specific T-cell responses. Three weeks after infusion, CD4+ and CD8+ T cells were 124 ± 54/cmm and 327 ± 42/cmm (versus 11 ± 4/cmm and 8 ± 4/cmm respectively, in patients receiving T-cell doses below 5 x 10 5 /Kg, P = 0.0007). Aspergillus, Candida, CMV, ADV, HSV, VZV, Toxoplasmaspecific CD4+ and CD8+ T-cell responses had recovered to frequencies within the normal ranges while they were absent in patients who received T cell doses under 5 x 10 5 /Kg (P = 0.0002).

In conclusion, this study demonstrates the feasibility, safety and preliminary indications of efficacy of adoptive immunotherapy after photodynamic purging of alloreactive T cells in recipients of haploidentical stem cell transplantation. A larger study will evaluate the impact of these T-cell infusions on transplant related mortality and disease free survival.

Leukemic stem cells (LSC) are a self-renewing subpopulation of malignant cells which is thought to play a central role in the pathogenesis of acute leukemia. LSC are likely contributing to both disease initiation and relapse and therefore represent an important therapeutic target. Resistance of LCS to standard chemotherapy agents is an important consideration for the development of new therapies. In this study, we asked whether LSC of acute myeloid leukemia (AML) can be recognized and eliminated by the anti-tumor function of natural killer (NK) cells. Using CD45dimCD34+CD38-as a phenotypic characteristic of the LSC population in AML, we purified these cells from peripheral blood of patients with de novo AML and examined their sensitivity to alloreactive NK cells. For this purpose, the recently described NK cell lines with single KIR specificities and mismatched with respect to HLA-class I allotype of target leukemic cells (Diermayr et al, Blood 2007 online) were employed. Using the hematopoietic colony forming unit (CFU) assay in methylcellulose, we demonstrated that purified LSC gave rise to hematopoietic colonies which were maintained upon a serial passage. When LSC were preincubated with alloreactive NK cells the growth of AML CFU was significantly reduced. Next, we demonstrated that CD45dimCD34+CD38-AML cells do not express the cell surface ligands, MIC and ULBP, for the major NK cell activating receptor NKG2D. Incubation of purified CD45dimCD34+CD38-cells with histone deacetylase (HDAC) inhibitor Valproic Acid (VA) for 2 days, resulted in an up to 2fold upregulation of MIC or ULBP ligands in 8/12 tested AML samples. The VA treatment had no impact on AML CFU formation in the absence of NK cells, but abrogated the clonogenic growth when applied together with alloreactive NK cells. Importantly, the effect of VA was limited to LSC since normal hematopoietic stem cells were unaffected, both with respect to expression of NKG2D ligands as well as the colonyforming efficiency. Taken together, our results indicate that AML LSC are susceptible to NK cell-mediated cytotoxicity and that expression of NKG2D ligands makes LSC more accessible to NK cells. We propose that adoptive transfer of alloreactive NK cells in combination with pharmacological use of HDAC inhibitors merit evaluation as novel approaches to prevent relapses of AML with NK cell immunotherapy. These approaches are now evaluated in an in vivo model of human AML transplanted into NOD/SCID mice.

Survival in patients receiving a transplant from an HLAmismatched donor (at 10/10 alleles) can be significantly improved by selecting a donor mismatched for HLA-DPB1 B. Shaw* (1, 2) , N. Mayor (2) Recipient/donor HLA matching is an important determinant of outcome in transplantation using volunteer unrelated donor (VUD). There is evidence that matching for HLA-A, -B, -C, -DRB1, -DQB1 results in a beneficial outcome. The impact of matching for HLA-DPB1 is more controversial and few studies have considered the additive effect of --DPB1. We investigated the outcome, dependent on the degree of HLA matching at 12 alleles, in 488 recipients of VUD transplants for leukaemia. The patients had AML (188), ALL (157) and CML (143). The majority of patients had T cell depletion as GvHD prophylaxis. 318 patient/donor pairs were matched for 10/10 alleles. Of these, 89 were matched in addition for DPB1. 118 pairs had a single allele mismatch (9/10). Of these, 30 were matched and 88 were mismatched for DPB1. 52 pairs had multiple mismatches (≥ 2), in 14 DPB1 was matched. We found a significant difference in overall survival when comparing these groups (fig 1, p=0 .006). Survival was not significantly different in the 10/10 matched transplants dependant on DPB1 matching (p=0.13). However, there was a significant difference in survival dependant on DPB1 matching within the HLA mismatched group (n=170) (4 years: 39% DPB1 mismatched compared 21% DPB1 matched, p=0.008). In particular, in pairs with a single HLA mismatch at -A, -B, -C, -DRB1, -DQB1, the presence of a DPB1 mismatch improved overall survival (4 years: 43%) compared to DPB1 match (25%, p=0.05). In multivariate analysis including disease, stage, patient/donor age, patient CMV status and gender, the significant survival benefit of DPB1 matching persisted (OR 0.502; 95% CI 0.32, 0.78; p=0.002). This effect appeared to be mediated both by a decrease in relapse in the DPB1 mismatched pairs (4 yrs: 47% compared to 87% in DPB1 matched pairs, p=0.004) and a decrease in TRM (4 yrs: 37% compared to 64% in DPB1 matched pairs, p=0.006). No differences in acute or chronic GvHD were seen. We conclude that in patients who receive HLA-mismatched grafts (at HLA-A, -B, -C, -DRB1, -DQB1), a donor who is DPB1 mismatched in addition results in a superior outcome compared to one who is DPB1 matched. As physicians frequently have a choice between mismatched donors, we believe that selection by DPB1 within this group will be possible and beneficial. Additionally, DPB1 is the only HLA allele for which matching status impacts on disease relapse. We speculate that this may suggest a difference in the functional mechanisms of DPB1.

The CXCL16/CXCR6 chemokine pathway specifically targets alloreactive T-cells to the bone marrow of mice with leukaemia R. Van der Voort, V. Verweij, F. Maas, J. Vos, M. Philippens, T. De Witte, H. Dolstra* UMC St Radboud (Nijmegen, NL) Allogeneic stem cell transplantation is an effective treatment for patients with leukemia. This therapeutic effectiveness is largely attributed to the graft-versus-tumor (GVT) response during which alloreactive donor T cells eradicate minor histocompatibility antigen (MiHA)-expressing tumor cells in the bone marrow (BM). Unfortunately, subsets of alloreactive T cells recognize MiHA in healthy tissues, such as the gut, skin and liver, resulting in graft-versus-host disease (GVHD). Thus, there is a strong need to separate GVT responses from GVHD. An appealing strategy to achieve this goal would be to increase the migration of alloreactive T cells to the tumorcontaining BM and/or block the infiltration of GVHD-prone organs. Although the homing receptors for migration into the gut and skin are extensively studied, those involved in T cell trafficking to the BM are largely unknown. Here, we characterized the pathway that alloreactive T cells exploit to infiltrate leukemia-containing BM. T cells were isolated from BM and spleen (control organ) from mice with or without acute myeloid leukemia (AML) that had previously received an allogeneic BM-transplant (BMT). In addition, we used naïve mice as controls. The expression of activation markers and a panel of 14 putative homing receptors was determined by flow cytometry, and the level of several chemokines by real-time PCR and ELISA. Migratory behavior was analyzed by Transwell migration assays and flow chamber assays. Where BM of naïve and BMT mice contained few T cells, AMLcontaining BM was heavily infiltrated with CD4+ and CD8+ T cells with an effector/effector memory phenotype. Interestingly, BM of mice with leukemia predominantly recruited CD4+ and CD8+ T cells expressing the chemokine receptor CXCR6. In contrast, the number of CXCR6+ T cells in BM from naïve or BMT mice without AML, and in all spleens was generally low. Furthermore, the ligand for CXCR6, CXCL16, was expressed by activated macrophages and dendritic cells and its expression level increased in BM of mice with AML as compared to controls. Finally, we show that CXCR6+ T cells are strongly attracted by soluble CXCL16 and show increased binding to the endothelial adhesion molecule VCAM-1 upon stimulation with CXCL16. In conclusion, our data suggest that the CXCL16/CXCR6 axis is a novel pathway that specifically regulates the recruitment of alloreactive T cells into leukemia-containing BM.

The PD-1/PD-L1 co-inhibitory pathway is involved in functional impairment of alloreactive CD8+ T-cell responses against leukaemia W. Norde (1) , F. Maas (1) , H. Fredrix (1) , A. Schattenberg (1), M. Kester (2) , F. Falkenburg (2) , R. Van der Voort (1) 

Donor lymphocyte infusion (DLI) following allogeneic stem cell transplantation (SCT) is a potent treatment for hematological malignancies. The effectiveness is attributed to the graftversus-tumor (GVT) response, during which donor CD8+ T cells become activated by minor histocompatibility antigens (MiHA). Consequently, these alloreactive CD8+ T cells clonally expand, acquire effector functions and kill MiHApositive malignant cells. After the response, most CD8+ T cells die through apoptosis, while a small population remains to form a pool of long-lived memory cells. Unfortunately, tumor relapses still occur despite the presence of MiHA-specific memory T cells in the patient for many years, suggesting that these memory T cells become impaired over time contributing to tumor evasion. Recently, a crucial role for the co-inhibitory PD-1/PD-L pathway in inhibiting the function of virus-specific CD8+ T cells was demonstrated in chronic viral infections. Here, we investigated the role of the PD-1/PD-L pathway on the functionality of MiHA-specific CD8+ T cells. Initially, we analyzed expression of activation and co-signaling molecules on MiHA-specific CTL following restimulation with irradiated MiHA-positive EBV-LCL plus IL-2. We observed that PD-1 expression was upregulated on MiHA-specific CTL within 24 hours after stimulation. In addition, we analyzed MiHA-specific CD8+ T cells from leukemia patients after DLI. We observed increased levels of PD-1 on MiHA tetramer-positive CD8+ T cells, compared to tetramer-negative CD8+ T cell subsets in the same patient. Knowing that sustained expression of PD-1 is associated with diminished proliferative capacity of virusspecific CD8+ T cells, we determined whether blocking the ligation of PD-1 could increase the capacity of MiHA-specific CD8+ T cells to proliferate in vitro. Interestingly, we revealed that the addition of anti-PD-L1 antibody augments the proliferation of MiHA-specific CD8+ memory T cells. Notably, addition of peptide alone or in combination with an irrelevant antibody did not induce proliferation of these cells. Finally, we observed that primary leukemia cells and various hematological cancer cell lines express the PD-1-ligand PD-L1 following stimulation with IFNg. These data suggest that PD-1/PD-L1 interactions in the tumor environment may impair MiHA-specific CD8+ T cell function, and intervening with this co-inhibitory pathway may result in improved GVT immunity after allogeneic SCT.

The graft versus-leukemia-effect in B-cell chronic lymphocytic leukemia (B-CLL) appears to be less efficient compared to myeloid leukemias. Since the aberrant coexpression of CD5 in association with CD19 is a hallmark of B-CLL, we asked if both molecules might serve as a combinatorial T-cell target. B-CLL cells as third party in mixed-lymphocyte reactions (MLR) induced T-cell anergy. We used recombinant proteins comprising a single chain (sc) FvCD5 antibody (Ab) fused to human interleukin (IL)-2 via the hinge region of human immunoglobulin (Ig)G1 in conjunction with a bispecific S27 CD3x19 Ab (bsAb; clone OKT3xHD37) to saturate CD19 and CD5 binding sites on B-CLL cells. Consistent with the absent immunogenic potential, B-CLL cells pretreated with scFvCD5 Ab, native IL-2 and CD3x19 bsAb were not recognized by allogeneic resting T cells. In contrast, B-CLL coated with the dimeric scFvCD5-IL-2 and CD3x19 bsAb induced profound stimulator cell dependent allogeneic and autologous T-cell proliferation. In addition, T cells from fresh peripheral blood mononuclear cell samples from B-CLL patients (n=4) pretreated with dimeric scFvCD5-IL-2 and cultured in the presence of CD3x19 bsAb expanded by 2-3 log within 7 days. Notably, addition of soluble exogenous IL-2 to both types of culture remained inferior compared to the CD5-targeted IL-2 delivery. T-cell proliferation and expansion also resulted in cell-mediated cytotoxicity. In conclusion, a dual targeting approach using aberrantly expressed tumor cell surface molecules for membrane delivery of dimeric IL-2 molecules using a fusion cytokine in conjunction with a bispecific antibody as shown for B-CLL represents a strategy to reverse tumor-induced T-cell anergy. Hurler's syndrome (HS), the most severe form of mucopolysaccharidosis type-I causes progressive deterioration of the central nervous system and death in childhood. Allogeneic-stem cell transplantation (SCT) before the age of two years halts disease progression and prolongs life. Graft-failure and mixed-chimerism (40-50%) limit the success of SCT for HS. Unrelated-cord blood transplants (UCBT) are suggested to be a good alternative option for bone marrow, however, little is known about risk factors for outcomes after UCBT for this disease. We have analyzed 93 HS children that received an UCBT from 1995 to 2007 and were reported to EUROCORD or transplanted at Duke University. Median age at UCBT was 1,3 (0,2-4) yrs, and median follow up was 24 (3-140) mths. The donor was HLAidentical (HLA-A and B by low resolution and HLA-DRB1 by high-resolution) in 13 cases (16%) and incompatible in 67 cases (84%: most with 1 (54%) and 2(26%) HLA disparities). The median nucleated cell dose/kg and CD34+/kg at infusion were respectively 7.2 (2-22)x10 7 and 2,3 (0,5-17)x10 5 . With the exception of 5 patients, all received a Busulfan/Cyclophosphamide (+Fludarabine: 6) regimen. All patients received ATG or Campath (4). Results: Median days to neutrophil and platelet recovery were 22 (10-46) and 35 (13-82) days, respectively. Mixedchimerism was found in only 4%. All patients had normal enzyme levels after engraftment. In multivariate analyses for neutrophil recovery, a CD34±dose of >2.3x10 5 /kg (HR=2.0; p= 0.015) was associated with increased probability of recovery.

Acute-GvHD (grade II-IV) was observed in 27%, while chronic-GvHD was seen in 10% at 2 years. Two years overall survival (OS) and disease/event free survival were 78% and 70%, respectively. For 2 years OS, time from diagnosis to UCBT more than 6 months was associated with increased risk of death (6% for those children transplanted earlier and 30% for those transplanted later: p=0,04). In conclusion, outcomes following UCBT for Hurler syndrome are encouraging. UCB appears to be a good alternative allogeneic stem cell source to transplant children with HS. Earlier transplantation and higher cell dose are associated with better outcomes after UCBT for HS patients. Introduction: Transplant for MPSIH has been associated with high rates of morbidity, mortality and graft failure. Some of the difficulty of the transplant reflects pre-transplant co-morbidities including significant cardiac dysfunction, respiratory including upper airway compromise and visceral organ enlargement. In recent pharmacological enzyme replacement therapy (ERT) has been available that will ameliorate many of these comorbidities. Manchester protocol. We give 12 weekly doses of ERT prior to conditioning and continue until donor cell engraftment. Patients and Methods. We have performed 71 transplants in 54 patients in Manchester. 19 of these patients have received ERT with transplant. In 15 of these 20 there has been full intensity conditioning with pK guided oral (HD) busulfan (40mgs/m²/dose, n=7) or IV busulfan with pK monitoring but not adjustment. Results: Of these 15 patients 7 received oral busulfan and the most recent 8 have received IV busulfan. Of these 15 patients 14 are alive with full donor cell engraftment. One patient experienced primary graft failure following a cord transplant and was successfully transplanted with a second cord and despite achieving full donor chimerism died of adenovirus 18 months after the graft. Donor source for these 15 patients was MUD cord in 9, other MUD in 2 (PBSC donations) and matched family donors in 4 (PBSC in 2). EFS survival for this group is therefore 93% with a median follow up of 21 months. Our engrafted survival rates for previous patients treated on previous protocols are 68%, including second grafts to maintain donor chimerism. 4 previous patients receiving ERT and reduced intesity grafts experienced 3 rejections. Administered busulfan was significantly higher in those receiving HD busulfan -they received a mean of 26mgs/kg compared with the previous dose of 20mgs/kg -(p<0.001, students t-test). Discussion: Manchester is the largest centre for MPS transplants in Europe and this is the largest collection of patients treated with both ERT and SCT. There were concerns that pre-transplant ERT would increase rates of graft rejection. These data refute those concerns. We believe that the improving results reflect reduced patient co-morbidity after ERT, improved delivery of full intensity chemotherapy with busulfan and better donor matching including use of unrelated cord blood as a donor source.

S28 O180 Haematopoeitic stem cell transplantation for chronic granulomatous disease -a single-centre experience E. Soncini* (1), M. Slatter (1) , L. Jones (1), S. Hughes (1), T. Flood (1) , D. Barge (2) , G. Spickett (2) 

Chronic Granulomatous Disease (CGD) is a primary immunodeficiency affecting phagocytes; mutations in genes coding for subunits of nicotinamide adenine dinucleotide phosphate cause failure of oxygen metabolite generation, resulting in impaired microbial killing, susceptibility to bacterial & fungal infections and lung & gut inflammation. Antibacterial & antifungal prophylaxis with cotrimoxazole & itraconazole improve short and medium term survival but do not cure CGD; patient quality of life is burdened by frequent hospitalizations, recurrent diarrhoea, inflammatory lung damage, failure to thrive and a 50% risk of death by the third decade of life. Haematopoietic stem cell transplantation (HSCT) can cure the disease. We report our single centre experience. Between 1998 and 2007, 20 patients underwent HSCT for CGD in Newcastle. A retrospective analysis of medical records to November 2007 reviewed age at HSCT, donor type, HLA matching, conditioning, immuno-reconstitution, significant post-HSCT complications, growth, lung function and outcome. 15 received Bu/Cy, 3 Flu/Melph, 1 Flu/Bu, 1 Bu/Mel conditioning. 10 had sibling, 10 unrelated donors. 17 had 10/10 HLA-match, 2 9/10 HLA-match (A, C mismatch), 1 8/10 HLA-match (DR & DQ, HvG direction only). 15 received marrow, 3 PBSC and 2 cord blood transplants. 18 of 20 (90%) patients were alive with functioning neutrophils with follow up 0.25-9.25 years. 3/4 with significant inflammatory lung disease had an improvement in lung function. All 9 patients with colitis had resolution of symptoms, with impressive catch-up growth. 1 patient had cytopenias secondary to post HSCT-CMV infection, there were no other significant infections, only 2 had significant GvHD, 1 liver that resolved with treatment and 1 gut GvHD that resolved with prolonged steroid treatment, with resulting osteoporosis and avascular hip necrosis. 2 patients died, both of complications relating to active fungal infection. 13/18 survivors are off prophylactic medication with excellent life quality. HSCT is curative for patients with CGDcomplications are less frequent and outcome better in those without pre-existing infection or inflammation. HSCT should be considered early in the treatment of CGD, particularly if a suitable donor is available.

Long-term outcome and quality of life after haematopoetic stem cell transplantation in osteopetrosis: a single-centre experience D. Moshous (1)

Background: Malignant infantile Osteopetrosis (MIOP) is caused by defective osteoclast function. Impaired bone resorption is leading to increased bone density resulting in progressive marrow failure and extramedullary haematopoiesis. The abnormal bone remodelling generates also compression of cranial nerves, especially the optical nerve, causing various degrees of visual impairment up to complete blindness, starting early in life. Some patients present also neurological deterioration which is not clearly explained by the pathophysiology of MIOP. The genetic basis of OP is heterogeneous, until now four genes with autosomal recessive inheritance have been identified: TCIRG1, CLCN7, OSTM1 and RANK-L, but there seem to be also autosomal dominant forms. In the absence of treatment, children with MIOP die early in life often because of bleeding or infection. The only curative treatment available consists in Haematopoetic Stem Cell Transplantation (HSCT). Patients and Methods: We report on 40 HSCT which were performed in 35 patients (17 females and 18 males) in our unit between 1984 and 2007. Stem cell sources were bone marrow, peripheral blood stem cells and one unrelated cord blood unit. The donors were haploidentical related donors (26/40), HLA-identical siblings (9/40), related matched donors (2/40), MUD (1/40) and unrelated cord blood (1/40). The majority of patients received HSCT before the age of three months (16 days to 26 months), with a mean age at first transplant of 5,5 months. Two patients have been transplanted at an older age, at 3 years 4 months and 8 years 5 months respectively. Conditioning consisted mostly in Busulfan and Cyclophosphamid based regimens, in 5 cases Fludarabine was added, ATG was used in 12 cases. The injected cell number varied from 0,2 -19,4 x 109 nucleated cells/kg, corresponding to 2,3 -73,2 x 106 CD 34+ cells. Outcome: For two patients the follow up is still too short to conclude. Out of 33 evaluable patients 16 survived (overall survival: 49%). The causes of death were interstitial pneumonia in 4 patients, Veno-occlusive disease in 2 patients, multiorgane failure in 2 patients, rejection in 4 patients, transplanted related toxicity in 3 patients, bleeding in 1 and infection in 1 patient. Survival was better in the setting of HLAmatched HSCT 8/12 (66,7%) versus HLA-mismatched HSCT 8/21 (38%. Three patients received 2, one patient 3 HSCT. Detailed data especially on long-term neurological development will be provided. (2) ). There was no graft rejection; mixed donor chimerism occurred in 36%. The incidence of an acute GvHD ≥ II was 19 %, that of a chronic GvHD II 11 % with no difference between MRD and MUD. cGvHD was associated with more than expected MRI changes 2 years after HSCT (p < 0.05). Moreover, patients with unexpected disease progression often displayed poor immunologic reconstitution and late viral problems. 14 of 18 pts. (= 78 %) in group II/III remained neurologically stable with good quality of life; 12 of these continued to visit a regular school. All pts. in group I deteriorated after HSCT; however, 4 out of 9 pts. stabilized and full dementia could be prevented. Conlusion: The above results confirm the role of a favorable disease stage for HSCT outcome, as already shown (C. Peters, Blood, 2004) . In addition, the data demonstrate the importance of transplant-associated factors (e.g. cGvHD). In our view, the superior neurologic outcome after MRD-BMT allows a less restrictive indication for HSCT in more advanced CCALD patients.

J.-S. Kühl* (1), G. Strauß (1), B. Weschke (1), W. Köhler

First successful bone marrow transplantation for X-linked chronic granulomatous disease using a sibling donor selected after preimplantation female sexing and HLAmatching J. Reichenbach* (1), H. Van de Velde (2), M. De Rycke (2), C. Staessen (2) , P. Platteau (2), P. Baetens (2), T. Güngör (1), (1), I. Libaers (2) (1

Objective: Allogeneic hematopoietic stem cell transplantation (HSCT) from an HLA-identical donor is currently the only proven curative treatment for chronic granulomatous disease (CGD). HSCT with alternative donors is associated with higher morbidity and mortality. The objective was therefore to perform in vitro fertilization (IVF) and preimplantation HLAmatching combined with female sexing for HSCT in X-linked CGD. Methods: IVF followed by preimplantation genetic diagnosis (PGD) was used to identify a female HLA-genoidentical embryo in a family who needed a suitable donor for their boy affected with CGD. Two PGD cycles were performed. Results: In the second PGD cycle two HLA-genoidentical female embryos were transferred and a pregnancy was obtained. The mother had a normal delivery of a healthy girl at term. Conventional HSCT had to be performed, due to insufficient cell numbers in the cord blood source, at age 12 months of the donor and age 5 8/12 years of the recipient and resulted in complete stable donor chimerism and immunological reconstitution up to 17 months post HSCT. Conclusion: HSCT after IVF and combined female sexing and HLA matching offers a new therapeutic option for patients with X-linked primary immunodeficiency such as CGD needing HSCT but lacking an HLA-genoidentical donor.

A multicentric comparative analysis of outcomes of HLA identical related cord blood and bone marrow transplantation in patients with beta-thalassemia or sickle cell disease N. Kabbara* (1), F. Locatelli (2) , V. Rocha (1) Most patients with beta thalassemia major (TM) or sickle cell disease (SCD) can be cured by hematopoietic stem cell transplantation (HSCT) from either cord blood (CB) or bone marrow (BM). One advantage of CB is the absence of risk associated with donation. In order to compare outcomes after HSCT with CB or BM, we studied 388 patients with TM or SCD who received HLA identical sibling CB (n=72) or BM (n=316) allografts between 1994 and 2005. In order to avoid center and period effect, only centers that performed both types of HSCT during the same period were included. We compared the incidence of hematopoietic recovery, acute and chronic graft-versus-host disease (GvHD), disease-free survival (DFS) and overall survival (OS) after CB and BM transplantation. Compared to BM, CB recipients were significantly younger (median of 6.2 y versus 7.2 y), smaller (19 kg vs 24 kg), and were transplanted more recently (in 2001 vs 1998) . The BM group consisted of 127 (40%) SCD and 189 (60%) TM patients, and the CB group, 26 (36%) SCD and 46 (64%) TM patients. The indications for transplantation in SCD were not statistically different between CB and BM groups. More TM patients belonging to Pesaro II-III risk classes received BM (65%) compared to CB (36%) (p=0.004). There were also differences in the conditioning regimen (more frequent use of ATG/ALG in the BM group and of Fludarabine and Thiotepa in the CB group) and GVHD prophylaxis (more methotrexate-containing therapy in the BM group compared to the CB group). In addition, the nucleated cell content was 10 times higher in BM compared to CB. The table below shows the non-adjusted univariate analysis for outcomes for all patients according to the stem cell source used. In TM patients, the 5 year-DFS rates were 87% and 83% for BM and CB recipients, respectively, and in SCD patients, 92% and 85%, respectively. In a multivariate analysis adjusted for age and type of hemoglobinopathy, DFS was not statistically different between CB and BM recipients (RR=1.4, p=0.34). In conclusion, patients with TM or SCD had excellent outcomes after HSCT whether they received stem cells of CB or of BM from an HLA identical sibling. These results strongly suggest that CB transplantation from HLA identical siblings should be pursued when possible to avoid the discomfort and risks of a bone marrow harvest. S30 O185 Communication biases during informed consent for unrelated bone marrow transplantation in adult thalassaemia patients G. Caocci* (1), S. Pisu (2) (1), G. La Nasa (2) (1)R. Binaghi Hospital -ASL8 (Cagliari, IT); (2)Cagliari University (Cagliari, IT); (3)GIMEMA Data Center (Rome, IT); (4)Oxford University (Oxford, UK) Although there are numerous guidelines to evidence-based medicine, few explain how to build the information into patient oriented decision-making. Several factors may hamper physician-patient communication and compromise the informed consent process. In an attempt to eliminate some of the barriers to understanding, we investigated the main factors of communication between physicians and adult thalassemia patients transplanted from an unrelated donor Twenty-five transplanted thalassemia patients and 12 physicians were given a questionnaire to investigate the communicated, perceived and recalled risk for mortality and graft-versus-host disease (GvHD), the acceptable risk percentage, besides the motivation and external influences underlying the choice of undergoing HSCT. The risks of dying or GvHD perceived by the patients were significantly lower than those communicated by the physicians (p=.002, p<.001, respectively). Also the perception of severe GvHD as a life-threatening condition was much lower (p=.004). Younger patients perceived a significantly higher risk of dying than older patients (p=.003). Females perceived a significantly higher risk of developing GvHD or dying than males (p<.05 in both cases). The median percentage communicated for the risk of dying was significantly higher than the percentage remembered by the patients (30% vs 20%; p=.003). The median percentage considered to be acceptable for the risk of dying was significantly higher in the patients compared to the physicians (30% vs 20%; p=.008). Hence, the balance between the risks and the benefits of the transplantation procedure had not been fully understood by the patients. Therefore, it is important that physicians make every effort to overcome communication bias, heuristics and distorted processes of remembering or understanding that could possibly influence the informed consent and decisionmaking process. In adults with AML (n=24578), 43% received an HLA identical family HSCT, 35% an autograft, 14% an HLA matched unrelated HSCT, 8% an HLA mismatched HSCT(3% haplo, 4% HLA unrelated and 1% cord blood). In adults with ALL (n=10932), 44% received an HLA identical family HSCT, 23% an autograft, 21% an HLA matched unrelated HSCT, 12% an HLA mismatched HSCT (4% haplo, 6% HLA unrelated and 2% cord blood). For adults in first CR with ALL or AML given a HLA identical sibling donor, 5-years overall survival (OS) was 57%; it was 49 % for those given an autograft and 49% for those patients given a HLA matched unrelated transplant. For adults with AML given an unrelated cord blood or a haplo-transplant in first CR, overall survival at 2 years were 62% and 48% respectively. Many studies have been performed and published on behalf of the ALWP with the collaboration of many EBMT centres. Thanks to the EBMT centres, physicians and data managers, we have been able to collect specific information for those studies. Therefore, in 2007, 5 studies and one editorial have been published in the most important journals in the field of medicine, hematology and transplantation, 13 retrospectives and 2 prospective studies are on going and 3 of 5 abstracts were oral presentation at the last ASH meeting. Moreover, the establishment of 5 subcommittees within the ALWP, in the different fields of transplantation has brought enthusiasm among participants and increased the number of participants to up to 50 in the ALWP meetings. We would like to thank W Arcese, B Labar and O Ottman for their input as heads of alternative donor, developing centres and Phi+ALL subcommittees, respectively. They have accomplished two years as subcommittee's heads. We would like to welcome A Nagler, S Giebel and J Esteve as new heads of alternative donor, developing centres and molecular markers subcommittees, respectively. I take also this opportunity to thank M Mohty and C Schmid for their activity in the ALWP. We hope that the enthusiasm inside the ALWP can motivate young physicians and more EBMT centres to join us with their ideas and input to perform important studies for the EBMT community and importantly to continue improving the field of transplantation for patients in need.

The main objective of this new subcommittee is the analysis of the impact of main molecular markers on the outcome of stem-cell transplantation. Several molecular lesions described in recent years are contributing to a biological characterization of acute leukemia and provide relevant prognostic information. In this regard, mutations of FLT-3, nucleophosmin (NPM1) or CEBPA genes, frequently found in the large subgroup of AML with normal karyotype, arise as the most relevant prognostic factors in this subset of patients. Nonetheless, the precise role of different transplant modalities for each of these AML categories remains to be defined. Thus, a recent analysis performed within the ALWP confirmed a higher risk of relapse associated with FLT-3 ITD after myeloablative allotransplant transplant from an HLA-identical sibling, although the procedure seemed to benefit a significant proportion of FLT-3 ITD AML patients transplanted in first CR, according to the survival plateau observed. Nonetheless, the optimal transplant policy in these molecularly-defined high-risk patients is not firmly established. The SC is also aimed to investigate the results of transplant in patients with rare subtypes, such as AML with translocation t(6;9)/DEK-CAN or AML with 3q26/EVI1 rearrangement. Given the low frequency of these entities, only those studies performed in large cooperative groups have the potential to elucidate the role of different transplant strategies for these specific subtypes. Another objective of the SC is the design of biologically oriented studies analysis on the outcome of transplant in adult ALL, which might help to clarify the controversial role of transplant in early phases of the disease. Thus, specific studies for wellrecognized prognostic categories of adult ALL such as MLLrearranged, hypodiploid varieties or results of transplant among different T-ALL subtypes can be sponsored by the ALWP. Finally, the emergence of molecularly targeted therapy such as tyrosine kinase inhibitors in Philadelphia positive ALL or conjugated antiCD33 monoclonal antibody in AML are changing the natural history of the disease and, therefore, might modify indications and results of transplant in patients treated with these agents.

Reduced-intensity conditioning M. Mohty* Institut Paoli-Calmettes (Nantes, FR) In the last decade, the so-called non-myeloablative or reduced-intensity conditioning (RIC) regimens for allogeneic stem cell transplantation (allo-SCT) have emerged as an attractive modality to decrease allo-SCT-related toxicities. Indeed, RIC allo-SCT represents an attempt to harness the well-documented immune graft-versus-tumor (GVT) effect while attempting to overcome toxicity. The work of different pioneering groups rapidly proved that this approach is feasible in several disease settings or patients' categories, and had the added benefit of expanding the transplant option to patients who are ineligible for myeloablative allo-SCT. Unfortunately, and despite several thousands of patients receiving RIC allo-SCT reported to international registries, the true value of RIC allo-SCT in the management of haematological malignancies is, as yet, difficult to delineate. Several reasons can help understanding these difficulties. A consistent definition of "non-myeloablative" or 'RIC' regimens is still lacking. The different RIC regimens comprise a continuum that overlaps with standard myeloablative regimens. Where in theory, reduction of the "inflammatory" component of the conditioning as in RIC regimens may lead to the reduction of the incidence of GVHD, GVHD remains a matter of concern after RIC allo-SCT, raising questions about the need for continuous immunosuppression and its corollary of long-term infectious complications. Also, the notion or definition of "ineligibility" for conventional standard allo-SCT is not clearly defined. The goal of the RIC subcommittee within the Acute Leukemia Working Party of EBMT is to address through retrospective and prospective studies, the specific role of RIC allo-SCT in net health outcomes that should include, in a specific disease setting, an analysis of disease-free survival and overall survival balanced against treatment-related toxicity, quality of life, complications and death.

Transfusion of donor lymphocytes (DLT) for treatment of leukaemia relapse after allogeneic stem cell transplantation (SCT) has been a milestone in the field of immunotherapy against malignant disease. It can be regarded as the proof of principle for the Graft-versus-Leukemia effect. However, in contrast to the impressive results obtained in CML, results in acute leukemias have been inferior, although there is a small subgroup of patients that obviously responded. The immunotherapy subcommittee of ALWP has set the goal to study the role of cellular immunotherapy more in detail with respect to different diseases as well as distinct variants of allogeneic SCT. Hence, a large retrospective analysis of DLT in relapsed AML has been performed, and similar investigations have been started in ALL. Further, the role of DLT after haploidentical and RIC transplants is in the focus of ongoing studies. Finally, the conditions and the results of prophylactic or preemptive DLT, given after SCT to chimeric patients in CR or in a minimal residual disease status, are currently investigated.

Use of HSCT for patients with acute leukemia is increasing in Europe. Therefore many EBMT centers have been established in all countries, mainly in the new EU countries. Although the number of transplants in Eastern countries is growing and the results are improving, the scientific activity appears still insufficient. Enhancement of the scientific collaboration in Eastern Europe is a principle goal of the Developing centres subcommittee. To reach this goal it is planned to create a panel of young investigators representing major centres from the region. This panel is expected to discuss current clinical practice as well as to share laboratory experience. On this background it will be able to propose and coordinate multicentre studies including: 1) prospective clinical trials, 2) retrospective analyses, 3) biological studies on HSCT for acute leukaemia. Other goals of this subcommittee is to evaluate outcomes of HSCT performed in Eastern countries as well as to develop studies and methods for analyzing the role of center effect on the HSCT outcomes. With these objectives two studies have been performed. In the first one, we have analyzed 640 patients (AML, 459; ALL, 181) treated with HLA matched related donor in first complete remission in 10 countries, from1990 to 2006. With this study we have observed that results of MRD-HSCT for acute leukemia in Eastern Europe improved over time, as a consequence of decreased NRM. Another study performed was the update of the Lancet paper which has demonstrated a center effect in major EBMT centers. We have updated this study and we were able to show a persistent centre effect but improvement of LFS of HLA identical HSCT for adults with AML in first CR over the period 1987-2005. Other study proposals have been discussed such as the use of economic and social markers in order to evaluate transplant outcomes in the different EU countries in collaboration with yearly survey of A Gratwohl.

Currently, we can find an alternative donor for almost all patients without an HLA identical sibling donor, such as HLA matched unrelated, haploidentical and cord blood unrelated donor. The objective of this subcommittee is to study the feasibility of those different strategies and their outcomes and compare their results. With the active collaboration with Eurocord, we are able to perform studies of unrelated cord blood transplantation. A recent study in collaboration with Eurocord, that will be presented during this meeting, has shown the impact of KIR ligand mismatching on decreasing relapse and improving LFS in UCBT recipients with acute leukemia. Also comparative studies of outcomes after UCBT and haplo have been performed in children and adults. One of the main objectives of this subcommittee is to improve the HLA data of the database and therefore be able to study the impact of specific HLA alleles on outcomes of HSCT recipients, such as the impact of permissive HLA-DPB1 alleles. With the collaboration with the immunobiology WP and donor registries, we hope to improve the data of HLA high resolution typing of unrelated donors. Other projects are i) to study the influence of gene polymorphisms with outcomes of alternative HSCT ii) to compare various conditioning regimens in this setting; iii) to target therapy with the novel drugs preand post-alternative transplants and iv) to compare immune reconstitution after MUD vs. Haplo vs. CBT.

Update of the registration study V. Rocha*, E. Polge, M. Arat, V. Koza, H. Wandt, T. Ruutu, A. Ferrant, G. Milone, P. DeFabritis, W. Arcese, L. Verdonck, A. Bosi, B. Allione, A.L. Herr, F. Pinto, C. Arrais, S. Nabhan, G. Socie, G. Dini, E. Gluckman, M 

The Registration study is a joint prospective, multi-center, non randomized trial of EBMT AL and PDs WPs and Eurocord. Primary aim is to evaluate the contribution of different strategies of treatment (Matched sibling donor;MSD and an alternative HSCT) for adults with AL. Secondary aim are to compare, in a first step, policy of each centre, feasibility of each transplant modality, time to transplant. In a second step, to compare in a intent to treat analysis LFS, RR and TRM according to different transplants strategies. The date of the registration of the patients is considered as the date of HLA typing test. For each patient, a specific questionnaire has to be completed specifying the initial strategies and each change in those strategies : at registration, after 3-6 months after the resgitration and , each 3 months during the first year after HSCT, and twice a year for the following 2 years. The intention to treat will be defined, for each patient by the choice of the treatment planned and reported by the each center and the real modality performed according to the last option, and donor availability. Between December 2003 and December 2006, 727 adults, enrolled by 34 EBMT Centrers, were registered for the study, 498 had AML, 216 ALL and 13 Biphenotipic AL. HLA typing was performed for 448 patients at diagnosis, 179 after first CR, 43 in primary refractory LA, and 53 in more advanced phase (18 in CR2 or more and 35 in relapse). A MSD was found for 283 patients. The choice registered for the remaining patients was: 215 alternative donor (search for UD, CB, or Haplo); autologous HSCT (88 pts). chemotherapy alone (141 pts).The registration of patients has been closed on December 2006.The follow up of patients will proceed for 3 years. The final analysis will be performed on December 2009.

A randomised phase III study comparing conventional chemotherapy to low dose total body irradiation-based conditioning and haematopoietic cell transplantation from unrelated donors as consolidation therapy for elderly (>60y) patients with AML in first complete remission D. Niederwieser* on behalf of OSHO /FHCRC/HOVON SAKK/Austrian/French and ALWP-EBMT Study objectives: Efficacy of allogeneic related and unrelated hematopoietic cell transplantation (HCT) after reduced intensity conditioning as a consolidation treatment for patients with AML in complete remission or refractory anemia with excess of blasts (RAEB) in comparison with a non-transplant approach. Patients with a matched sibling or with an unrelated donor, who have entered CR1, will be eligible for randomization in a 2 (transplantation):1(non-transplantation) fashion. Patients without a donor will receive post-remission therapy without transplantation. Blaise, F. Frassoni, M. Kuenz, B. Rio, N. Russel, P. Rebulla, G. Sanz, J. Garcia, J. Cornelissen, C. Navarette, D. Niederwieser, A. Nagler, G. Socié, A. Sureda, V 

Umbilical cord blood (UCB) cells from unrelated donors have several advantages as compared to other sources of stem cells for allogeneic use, such as prompt availability, decreased risk of graft-versus-host disease and easy collection with little risk to the mother or to the newborn. UCB has been shown to contain sufficient progenitor cells to provide durable engraftment. However, because of low infused cell doses, single UCB transplantation (UCBT) in larger children and in adults may be associated with delayed engraftment and increased risk of graft failure. Transplantation of double cord blood units (DUCBT) represents a strategy to overcome this limitation. The results of both pediatric and adult DUCBT studies suggest that this approach is safe, and is associated with higher engraftment rates and improved transplant outcome, in both the nonmyeloablative and myeloablative settings, as compared to single UCBT. Some preliminary data also indicates a reduction in relapse with double UCB transplantation. In this view, we will discuss a proposal of a randomized trial comparing single versus double UCBT for patients with hematological malignancies. A writing committee has been established to discuss all the aspects of this protocol that will be presented during the ALWP session. Overall the leukaemia free survivals (LFS) at five years for patients autografted in first (CR1) and second (CR2) remission have been respectively 49 and 36% in adults and 66 and 49% in children, with some improvement in outcome after 1994. Recent prospective trials from single institutions or national groups have confirmed a LFS of 50% at 5 years in adult patients autografted in CR1. Several randomized studies in the pre "High dose Ara-C" era had shown superior outcomes of ASCT over conventional chemotherapy. The only randomized study using HDARA-C in the chemotherapy arm (US intergroup study,P Cassileth NEJM 1998, M Slovak Blood 2000) indeed has shown better results for ASCT in good risk patients by cytogenetics, better results for HDARA-C in intermediate risk patients and better results for genoidentical allogeneic transplants in the high risk group. Numerous retrospective studies using the EBMT registry have shown equivalence for LFS and overall survival (OS) when comparing ASCT to allogeneic transplantation with unrelated donors and/or allogeneic transplants with reduced intensity conditioning (RIC), with a higher relapse incidence (RI) for ASCT but a reduced transplant related mortality (TRM). Despite this situation, the annual reports of EBMT activities indicate a clear drop in ASCT for AML associated to a steady increase of allogeneic transplant activity, using RIC and all cell sources including cord blood. The practice of ASCT itself has changed over the years: less than 15% of the patients still receive marrow as a source of stem cells and mobilization of stem cells in peripheral blood has become the S33 standard. Total Body irradiation (TBI) which has been shown to have the highest anti leukemic effect is used in only 15% of the patients for conditioning. In vivo purging consisting of chemotherapy consolidation courses given before mobilisation is not standardized, and higher CD34+ cell yields have been shown to contain leukemic cells and to be associated with a higher relapse incidence (N Feller, Leukemia 2003) . A recent retrospective survey of the AWLP-EBMT (Presented at this meeting) indeed indicates that the number of consolidation courses given before autografting with peripheral blood influences the outcome. In patients autografted with marrow, peripheral blood early (interval CR1-transplant<80 days) and peripheral blood later (>80 days) , the LFS at 3 years have been respectively 52,46 and 36%. The source of stem cells and the interval from CR1 to transplant have been significant in multivariate analyses. Another recent retrospective survey from ALWP-EBMT (NC Gorin, JCO, 2008 in press) has compared ASCT to allogeneic stem cell transplantation in patients with Core binding factor mutations (inv 16 and t(8;21)) and shown equivalence in outcome for patients transplanted in CR1 with LFS around 70% at 3 years. Overall the data accumulate to confirm previous and already old observations that ASCT remains an interesting alternative therapeutical approach for the treatment of : -Older patients -Patients considered for RIC and/or unrelated donors transplants, with BM,PB or cord blood -Chemosensitive AML (so called good risk patients) including thoses reaching CR1 rapidly (rapid remitters) and those carrying a CBF mutation. Since PB has replaced almost totally PB as source of stem cells, for practicality reasons, all effort should be made at increasing in vivo purging aggressively, with, as much as feasible, minimal residual disease monitoring by molecular biology. There is still a need for randomized studies in good and intermediate risk patients, while high risk patients should be allografted whenever possible. In the largest prospective mutlicentre study initiated by the EBMT and comparing Filgrastim-mobilized peripheral blood progenitor cell (PBPC) to bone marrow, patients transplanted with PBPC developed more often chronic GvHD. However outcomes, which where overall and disease free survival, relapse and transplant related mortality were similar in both cohorts. At the time of first publication we were aware that long term observation was needed to ultimately determine the role of both source of stem cells. Indeed, outcome differences due to difference in the composition of the graft might become apparent only years after transplantation. We therefore planed a follow-up study in order to asses the long term outcomes, general health status, social integration as well as the occurrence of malignant and non-malignant late effects in both treatment groups. The patient accrual of the initial study took place between February 1995 and September 1999, including 350 patients transplanted for leukaemia from their HLA-identical sibling. In September 2007 we sent out a questionnaire to all centres which had participated to the study. The questions included information on survival, cause of death, relapse, secondary malignancies, chronic GvHD and its treatment, the occurrence of non-malignant late effects such as hypothyroidism, bronchiolitis obliterans, sicca syndrome, blood counts at last follow-up as well as question on general health status and social integration. So far, 20 (48%) from the 42 centres responded, including 105 of 212 patients (49%) alive at 3-year follow-up. Ninety-eight of the 105 patients reported patients are alive, and 7 have died after the 3rd year of follow-up (GvHD n=4, relapse n=2, other n=1). Chronic GvHD is still present in 23 patients (22%), secondary malignancies occurred in 4 cases (5%), non-malignant complications in 24 (22%). The median Karnofski score as a measure of general health was > 90% (range 70-100). Eightyeight (83%) patients have returned to work or to school. During the Late Effect Working Party Meeting 2008 in Florence, Italy, we will provide preliminary data on the longterm outcome, and compare both patients groups, patients transplanted with PBPC to those who had received BM. 

This is retrospective multicenter analysis on the incidence and risk factors of cardiovascular events after allogeneic hematopoietic stem cell transplantation (HCT), in 548 long term survivors treated in 10 EBMT transplant centers. These patients received HCT between 1990 and 1995 and surviving ≥1 year after transplantation. The median age of the patients at last follow-up or at time of a vascular event was 35 years (range 3-72 years), and the median follow-up time 9 years (1-16 years). Twenty (3.6%) out of 548 patients developed an arterial event in at least one arterial territory. The cumulative incidence of first arterial event 15 years after HCT was 6% (95% CI, 3%-10%). The cumulative incidence for patients with a high global cardiovascular risk score, including arterial hypertension, diabetes, dyslipidemia, physical inactivity and smoking was 17%, as compared to 3% in those with a low risk score. In univariate analysis, older age at last follow-up, all cardiovascular risk factors taken individually, and acute GvHD were associated with a higher risk of a cardiovascular accident. In multivariate analysis age older than 30 years at last follow-up, and patients with a high global cardiovascular risk score had a 6.4-fold and 9.8-fold increase, respectively of the relative risk for an arterial event after HCT. Thus, long term survivors after allogeneic HCT are likely to develop cardiovascular risk factors, and present an increased risk for premature cardiovascular accidents after allogeneic HCT. 

Update on the Phase II/III study of the incidence and outcome of VOD with the prophylactic use of defibrotide in paediatric stem cell transplantation (VOD-DF Study) S. Corbacioglu* University Children's Hospital (Ulm, DE) Along with graft versus host disease (GVHD) and cytomegalovirus (CMV) infection, veno-occlusive disease (VOD) is one of the most frequently encountered serious complications after stem cell transplantation. The currently reported overall incidence of VOD ranges from 5% to more than 60% in children who have undergone stem cell transplantation. Defibrotide (DF) is a polydisperse oligonucleotide derived from porcine intestinal mucosa with antithrombotic and protective properties on the microvasculature but minimal hemorrhagic risk. In large, multicenter, international phase I/II trials targeting patients with severe VOD DF has emerged as a promising therapy for VOD. The VOD-DF Study is a prospective international multicentre phase II/III study with the aim to assess the beneficial effect of defibrotide on the incidence, morbidity and mortality of VOD in children at high risk to develop VOD after myeloablative stem cell transplantation (NIH Trial Number: NCT00272948, Eudract Number:2004-000592-33) . The study, co-sponsored by Gentium and the EBMT, is open for recruitment since January 2006 with currently 27 participating centres in 10 countries. The prospective recruitment period is 3 years with a sample size of 270 patients. The current recruitment status as of January 2008 is 220 patients with an average recruitment rate of 10 patients per month. In November 2007 the first data safety monitoring board (DSMB) meeting evaluated safety data and mortality rates of the first 120 study patients who completed 30 days of follow-up. As conclusion of this meeting the types of adverse events described were considered typical for this patient population with no unexpected toxicities. No significant difference was observed in the number of adverse events between the prophylaxis arm compared to the control arm. Therefore it was unanimously concluded that there were no safety considerations of concern arising from the defibrotide-treated patients on the prophylactic arm at the present time. For the determination of the finale sample size an interim analysis will be performed at 240 patients. The recruitment of this sample size should be completed by early this year. In the meantime the following centres either completed or will complete the initiation process in order to join the study: Ankara, Antalya, Istanbul, and Prague. The Registration study is a joint prospective, multi-center, non-randomized trial of EBMT AL and PDs WPs in cooperation with the Eurocord. Primary aim is to evaluate the contribution of different strategies of treatment (MSD and alternative HSCT) for children with AL. Secondary aim are to compare, in a first step, policy of each centre, feasibility of each transplant modality, time to transplant. In a second step LFS, according to different transplants strategies, RR, TRM will be compared. The date of the registration of the patients is considered as the date of HLA typing test. For each step, a specific questionnaire has to be completed: at registration, after 3-6 months, at HSCT; each 3 months during the first year after HSCT, and twice a year for the following 2 years, followup forms will be completed. The intention to treat will be defined, for each patient by the choice of the treatment reported to be planned by the centre in each questionnaire, the real modality performed according to the last option, and donor availability. Between December 2003 and November 2007, 271 children, enrolled by 24 EBMT Centres, were registered for the study. Children had ALL (at diagnosis or in CR1, 94 patients; at first relapse or in CR 2, 73 patients; 12 in more advanced disease), or AML (at diagnosis or in CR1, 71 patients; with refractory disease, at first relapse or in more advanced phase, 17 patients). A MSD was found for 83 children, The choice registered for the remaining 183 children was: chemotherapy alone (45 pts 

It is estimated that every year in Europe about 500-600 children become donors of hematopoietic progenitor cells for siblings undergoing allogeneic stem cell transplantation. The aim of the study is the analysis of donor safety and occurrence of early side effects related to bone marrow (BM) or peripheral blood stem cells (PBSC) collection in pediatric siblings qualified to be a donor. The study is based on questionnaires send to transplant centers. Following endpoints are analyzed: complications during anesthesia, complications related to BM collection, complications of catheter placement, complications related to apheresis, number of nights spent in hospital and psychological issues in donor immediately after stem cell collection and during one year follow-up. Since January 2007, 112 donors (median age 9 years, range 11 months -18 years) were registered. The source of stem cells was PBSC and bone marrow, in 43 and 69 patients, respectively. Preliminary data indicate that the rate of complications is very low, thus the procedure is safe for donors. Due to low patients accrual, retrospective donors registration over a period of 1 year before the start of the study is being proposed. The 6th Meeting of the EBMT Pediatric Diseases Working Party will be held June 2-4, 2008 in Poznan (Poland) and for the first time will be accompanied by the Meeting of EBMT Pediatric Nurses. The Second Announcement with detailed information concerning preliminary scientific programme, call for abstracts, registration, hotel accommodation, travel, important dates etc. is available on the Meeting web page www.bokiz.pl/EBMT-PDs-WP-2008, which is also accessible using the link from the EBMT PDs WP web page. Physicians Sessions (Monday and Tuesday, June 2-3), Nurses Sessions (Tuesday, June 3), and Joint Session (Wednesday, June 4) will create an excellent opportunity to summarize recent progress in the field of HSCT from pediatric point of view, to initiate new ideas, and to improve an understanding and collaboration between nurses and physicians involved in pediatric HSCT. Apart from that the Meeting will be a unique platform for initiation new trials, interaction, and cooperation. More than 30 outstanding speakers will present various aspects of topics of the Meetings. However, organizers expect also a lot of interesting original oral as well as poster presentations, and an exciting and stimulating discussion. Therefore, all participants are invited and encouraged to submit their own experimental and clinical results for presentation during the Meetings. The deadline for abstract submission is Monday, March 17, 2008. Early registration is possible until Monday, March 31, 2008. Twenty grants from the EBMT covering registration fees and local accommodations will be awarded by the Scientific Committee on a competitive basis to physicians or nurses (information concerning application for grants is available on the Meeting web page).

Pharmaceutical companies, manufactures of technical equipment and software as well as publishers are invited to display their products at the industrial exhibition which will part of the Meeting. Detailed information about Poznan, including information how to get to Poznan, is available on the web page www.city.poznan.pl. It is a real honor and a great pleasure to invite everybody interested and involved in pediatric HSCT to Poznan to attend the 6th Meeting of the EBMT Pediatric Diseases Working Party and 1st Meeting of the EBMT Pediatric Nurses, Poznan (Poland), June 2-4, 2008. 235 Why is important to built a paediatric nurse group inside PDs WP? S. Calza* (1), V. van de Crommert (2) (1)Children's Hospital (Genoa, IT); (2) Transplantation group initiated a prospective, randomized, parallel-group, open-label phase III, multicenter study, comparing VTD (arm A) with TD (arm B) for MM patients progressing or relapsing after autologous transplantation. Inclusion criteria were: patients in first relapse after at least one autologous transplantation, including those who may have received Velcade or Thalidomide before transplant. Exclusion criteria: subjects with severe neuropathy or non secretory MM. 366 patients will participate (183 in each arm). Primary study end point was time to progression. Secondary end points included toxicity, response rate, eventfree survival and overall survival. Treatment was scheduled as follows: Velcade 1.3 mg/m² will be given as an i.v bolus on Days 1, 4, 8 and 11 followed by a 10-Day rest period (days 12 to 21) for 8 cycles (6 months) and then on Days 1, 8, 15, 22 followed by a 20-Day rest period (days 23 to 42) for 4 cycles (6 months). In both arms, Thalidomide will be given at 200 mg/Day per os for one year and Dexamethasone 40 mg/Day per os four days every three weeks for one year. Thrombosis prophylaxis was strongly recommended as well as valacyclovir prophylaxis (in arm A) against reactivation of varicelle zoster virus. Patients reaching remission could proceed to a new stem cell harvest. However, transplantation, either autologous or allogeneic, could only be performed after achieving the one year treatment. Response was assessed by EBMT criteria, with additional category of nCR. Adverse events are graded by the NCI-CTCAE, Version 3.0. As of January 15, 2008, 118 patients entered the study. 79 in France (IFM 2005-04 study), 10 in Italy, 9 in Germany, 8 in Switzerland (a SAKK study), 5 in Belgium, 3 in Austria and the Czech republic, 1 in the UK. 118 are assessable: 74 males, 44 females; median age: 60 yrs (range35-72), number of autologous transplant: one: 38, two: 80. Of these patients, 60 were randomly assigned to receive VTD and 58 to receive TD. Treatment was discontinued in 11 patients. An interim toxicity analysis was planned to be performed when the first hundred patients had been included. This interim analysis is ongoing (January 2008). Other preliminary studies demonstrate that VTD is a highly active and relatively well tolerated regimen. The combination is used in the relapse setting, as well as first line, consolidation and maintenance. In this protocol, the starting doses of Velcade and Thalidomide are relatively high and the duration of treatment is long. We will assess the superiority of VTD over TD in the relapse setting. Protocol EU-DRACT number: 2005-001628-35.

Tyrosine kinase inhibitors (TKIs) are the accepted first line treatment of choice for the majority of patients with chronic myeloid leukaemia (CML) although allogeneic stem cell transplantation (SCT) remains the only potentially curative procedure. Although second generation TKIs have shown promising results, in young patients resistant or intolerant to Imatinib and with a low EBMT score allogeneic SCT is still considered the best second line therapy. However, despite its curative potential, a significant proportion of patients relapse following allogeneic SCT. The gold standard for the treatment of relapse following allogeneic SCT is the infusion of lymphocytes from the transplant donor (DLI). However, DLI may result in significant complications including graft versus host disease and bone marrow aplasia and using escalating doses protocols, responses may be protracted. There is consequently a need for an alternative treatment for patients that relapse following SCT, that is both efficacious, faster acting, easier to administer and safer. Dasatinib has been shown to be effective in treating patients that are resistant or intolerant to Imatinib and as a result constitutes a good candidate treatment option for Imatinib intolerant or resistant patients that relapse following SCT. The CML sub-committee has launched a prospective phase II to investigate the efficacy of Dasatinib in this setting. This study will concentrate on patients 18 years of age with Ph+ CML whose disease has relapsed after transplantation from an HLA-identical sibling or an HLA-matched unrelated donor and have not responded to withdrawal of immunosuppressive treatment where this is possible. Enrolled subjects will be commenced on Dasatinib 100mg QID and receive treatment for 12 months. Continuation of treatment beyond 12 months will be at the investigator's discretion. The primary end-point will be complete molecular remission at 12 months secondary end-points will include complete cytogenetic remission rates, overall survival and proportion of patients requiring DLI. Donor lymphocyte infusions will be administered to all patients in whom Dasatinib has been discontinued indefinitely or there is evidence of disease progression during Dasatinib therapy or there is evidence of disease relapse after initial response tom Dasatinib. The study will open in the summer of 2008 and will aim to recruit 50 patients in 3 years. Interested EBMT centres are invited to participate. (11) Allogeneic hematopoietic stem cell transplantation (HSCT) as first line therapy for patients with chronic myeloid leukaemia (CML) has been replaced by imatinib. The role as second line therapy in patients who failed imatinib treatment is a matter of debate. Second generation Tyrosine Kinase Inhibitors (TKI) have already proven their efficacy in this setting. Early transplant related mor¬tality of allogeneic HSCT is considered to be too high. However, transplant outcome of young CML patients with a low risk for transplant related mortality has not been analyzed recently. Method: In order to better counsel patients with a HLAidentical sibling confronted with this situation we performed a retrospective analysis of transplants reported to the EBMT be¬tween 2002 and 2005. We selected for patients who had a low risk (EBMT-) score for transplant related mortality and who were transplanted from an HLA identical sibling. We analysed the outcome of those only who were transplanted in 1st chronic phase and who received best current treatment, defined as standard conditioning, no T-cell depletion and bone marrow as stem cell source. Results: 214 patients (8% of all 2737 patients transplanted for CML in this time period) with a median follow up of 12 months (0-60 months) who fulfilled these criteria were identified. They were 46% males and 54% females with a medium age of 31 years (range 6 to 59 years). 21% (46 patients) were less than 20 years old and 20% were above the age of 40. The time interval from diagnosis to transplantation was less than 1 year in 86% of patients. About one third each had an EBMT risk score 0, 1 or 2. Data were obtained from 81 teams in 33 countries. The probability of survival at 5 years in a competing risk model was 88% (95% c.i. 83-93) with a cumulative incidence of death without relapse of 10% at 14 months and no additional death from transplant related mortality thereafter until 60 months of follow-up ( Figure 1 ). 5-years post HSCT, 27% of patients were estimated to be alive after relapse (and hence the relapse free survival was 61%). Conclusions: These results show the current transplant outcome which is achievable by selecting only patients with a low risk for transplant related mortality. In this context, the data shown is valid even without information on pre-and/or posttransplant therapy. Allogeneic HSCT is a valuable option as second line therapy after imatinib failure for CML patients with a low transplantation risk. 

T prolymphocytic leukemia (T-PLL) is a rare, aggressive neoplasia of T lymphoid lineage which is characterized by poor survival of less than one year. Incidental reports suggest that both autologous and allogeneic hematopoietic stem cell transplantation (HSCT) might be effective in this disease. A comprehensive retrospective analysis of 44 patients registered in the EBMT database will be a subject of independent presentation at this meeting. However, given the limitations of conventionally collected registry data (dubious follow-up information and extreme heterogeneity), and realizing the impossibility of performing an international formal prospective trial under the current regulatory framework, we developed a new concept of two complementary projects: The first is called "EBMT prospective observational audit on allogeneic and autologous transplantation in T-PLL" and means that transplant centers will be encouraged to register their patients with T-PLL very timely with the EBMT, followed by mandatory submission of EBMT MedB and follow-up forms. The second is the "EBMT/ELN frame of orientation for allogeneic and autologous transplantation in T-PLL". Its purpose is mainly to avoid transplants in situations where they are very unlikely to be successful and to avoid excess heterogeneity of eventual transplants performed, thereby facilitating scientific analysis. This expert opinion-based framework covers criteria for the diagnosis of T-PLL, transplant eligibility, pre-transplant remission induction strategies, remission requirements, timing of HSCT, donor compatibility criteria, conditioning, GVHD prophylaxis, and MRD monitoring. With these two complementary components it should be possible to largely improve the usual quality of registry-based data and to generate scientifically sound knowledge on HSCT in an orphan disease such as T-PLL.

Primary plasma cell leukaemia (PCL) is a rare variant of plasma cell dyscrasia, associated with poor prognosis, median survival 8-12 months, significantly shorter than for multiple myeloma. Treatment with alkylating agent therapy is ineffective though polychemotherapy may offer modest improvement in survival. Autologous transplantation is now widely used in the treatment of PCL and this report summarises the European Blood and Marrow Transplant (EBMT) experience. A retrospective study was carried out of 20844 patients with common type myeloma (58% IgG, 21% IgA and 19% light chain) and 272 patients with PCL undergoing first autologous transplant between 1980 and 2006. All patients were reported using MED-A (limited data set) or MED-B (extensive data set) forms and included in the study regardless of availability of complete data. Comparisons used the Chi-squared test for categorical data and Mann-Whitney test for continuous data. Overall Survival and Progression-Free Survival were calculated using the Kaplan-Meier method and comparisons made using the Log-Rank test. Relapse/Progression and Death without relapse or progression were computed by the proper non-parametric estimator for outcomes with competing risks and compared by the Gray test. There was no significant difference in age, gender, calcium or albumin at presentation. Haemoglobin was significantly lower (9g/dl v 11g/dl, P=0.000), creatinine significantly higher (122 micro mol/l v 92 micro mol/l, P=0.000) and B2 microglobulin significantly higher in the PCL group. There was no difference in graft type or use of total body irradiation but the PCL group was transplanted closer to diagnosis (6.0 v 7.7 months, P=0.000). While no significant difference in engraftment, PCL patients were more likely than myeloma patients to enter CR post-transplant. Overall survival for PCL patients was greatly inferior to myeloma patients -25.7 (CI 19.5-31.9, P=0.000) v 62.3 months (CI 60.4-64.3), attributable to response of short duration and increased relapse-related mortality. This is the largest reported study of PCL patients and suggests improved outcome with use of autologous transplantation. It is however dispiriting to note that outcome is greatly inferior to that in myeloma despite likely pre-selection for fitness of the PCL group. There is urgent need for collaborative study of alternative approaches including highly effective induction with novel agents and optimal stem cell transplant strategy. 356 myeloma patients from 22 centres that had undergone HLA typing were included in the trial. Study inclusion was at the time of conditioning for first autologous transplant at the achievement of a response status of at least stable disease after VAD-like induction treatment of previously untreated patients. Patients with an HLA-identical sibling were allocated to the auto+allo (AuAl)-arm (n=108) and patients without a matched sibling donor to the auto (Au)-arm (n=248); single or tandem autografting was optional. Conditioning for ASCT was melphalan 200 mg/m 2 , and for allo RIC was fludarabine 30 mg/m 2 x 3 plus TBI 2 Gy. The accrual period was from February 2001 to February 2005. The two treatment groups were well matched for the standard prognostic parameters such as beta-2-microglobulin, karyotype, gender, MM subtype, stage, albumin, creatinin, calcium and response status at transplantation. Median age at transplantation was significantly higher in the Au-arm (57 vs 53 years). The CR rate was 43% in the AuAl-arm and 40% in the Au-arm (p=0.49). Cumulative 24 months non-relapse-mortality was 11% in the AuAl-and 4% in the Au-arm (p=0.05). At 3 years after transplantation, there was no significant difference between the treatment arms with respect to OS (AuAl 67%, Au 70%), RFS (AuAl 46%, Au 46%) or relapse rate (AuAl 43%, Au 48%).. However, looking at the OS curve for all patients in the AuAl-arm, a survival plateau on the 60%-level seems to be emerging from 3 years and onwards. We conclude that no significant differences in outcome was observed in this early analysis, but longer follow-up is warranted before any definite conclusions can be drawn. Updated results will be presented.

Allogeneic stem cell transplantation (SCT) can cure patients with MDS or AML. The major disadvantage of allogeneic stem cell transplantation is the high treatment related mortality. Recently the introduction of dose-reduced conditiong followed by allogeneic stem cell transplantation has lowered the treatment related mortality in comparison to standard conditioning, but a prospective comparison between both approaches is lacking. The subcommittee MDS of the CLWP launched a multicenter, prospective phase III-study comparing dose-reduced versus standard conditioning followed by allogeneic stem cell transplantation from related or unrelated donors in patients with MDS or secondary AML. The primary endpoint is treatment related mortality at one year. The hypothesis is that a dose-reduced conditioning will reduce the non-relapse mortality from 40 % to 20 % at one year after allogeneic stem cell transplantation. A total of 160 patients is needed to achieve this goal. Patients should have MDS or sAML (less than 20% blasts) and should be eligible for standard and dose-reduced conditioning and aged 18 -60 years if donor is a HLA-matched unrelated donor (HLA-A, HLA-B, HLA-DRB1 and HLA-DQB1) (one mismatch allowed)and aged 18 -65 years if donor is a HLA-matched related donor ((HLA-A, HLA-B, HLA-DRB1 and HLA-DQB1) (one anti¬gen-mismatch allowed). The patient will be randomised between a dose reduced conditioning (Arm B) and a standard conditioning (Arm A). The standard conditioning (Arm A) consisted of busulfan (16 mg/kg BW orally or 12.8 mg/kg intravenously) and cyclophosphamide (120 mg/kg) and reduced conditioning consisted busulfan (8 mg/kg BW orally or 6.4 mg/kg intravenously) in combination with fludarabine (150 mg/m²).

So far the protocol is activated in Germany, The Netherlands, Russia and Italy. Detailed protocol as well as recruitment will be presented.

A randomised trial of rabbit anti-thymocyte globulin, given on day+7 after alternative donor transplants A. Bacigalupo*, F. Ciceri, P. Di Bartolomeo, T. Lamparelli, G. Milone on behalf of the GITMO Background: Transplant mortality (TRM) can be predicted by using laboratory values, on day+7 after an allogeneic hemopoietic stem cell transplant (HSCT) (Bone Marrow Transpl.2003; 32: 205) : increased mortality is mainly, but not exclusively, due to increased acute graft versus host disease (GvHD). In a pilot study, a low dose of rabbit anti-thymocyte globulin (ATG) given on day+7 reduced GvHD and TRM. Aim of the study: The aim of this study was to test in a propsective randomized trial, whether intervention on day+7 in patients with a high risk score would reduced the risk of TRM and GvHD. Patients: Eligible were 170 patients undergoing HSCT from family HLA mismatched (n=25), or unrelated donors (n=145). All patients received ATG 3.75 mg/kg x2 pre-transplant, were stratified for intermediate and high risk day+7 score, and were randomized to receive an additional dose of rATG (1,25 mg/kg day +7 and day+9) (n=84) or no additional treatment (n=86) (control arm). The two groups were balanced for age, disease phase, day+7 score, and donor/recipient sex mismatch. Results: The predictive value of day+7 score on TRM was confirmed in the control arm (18% vs 42% for intermediate and high risk patients, p=0.03), whereas in the day+7 ATG arm , there was no difference (29% vs 29%, p=1.0). TRM was overall reduced from 35% in the control arm to 29% in the ATG day+7 arm (p=0.37) : the difference was more pronounced in patients with early disease and high risk on day 7 (35% vs 15% p=0.08), and in HSCT with female donors in male recipients (65% vs 20%, p=0.02). Acute GvHD grade III-IV was reduced overall from 16% to 6% (p=0.04), and chronic GvHD was reduced 32% to 14% (p=0.02). Conclusions: We confirm that patients with different risk of TRM can be identified on day+7 after HSCT: In patients at greater risk of TRM, the administration of ATG on day+7 reduces GvHD and TRM. Additional intensified supportive care (including anti-infectious treatment ) may further reduce TRM in patients with a high risk score on day+7.

Our group reported a strong association of polymorphisms (SNPs) within the antibacterial defense receptor NOD2/CARD15 with severe GvHD following allogeneic SCT. However, functional studies explaining these effects were so far missing. Therefore, we analysed gastrointestinal biopsies from 11 controls, 6 patients (pts) prior to SCT and 56 pts following SCT. Biopsies post SCT were obtained at the time of first symptoms indicating gastrointestinal GvHD. Slides were evaluated microscopically for occurrence of apoptotic cells, loss of crypts and infiltration with lymphocytes, eosinophils and neutrophils. In addition, immunohistochemical analyses for CD4, CD8, CD68, MIB1 and CD25 expression were performed. All pts and donors were typed for presence or absence of NOD2/CARD15 SNPs. Semiquantitative histological results were compared with clinical parameters such as stage of GvHD, use of corticosteroids and NOD2/CARD15 status. Comparison of controls with pts post-transplant showed a significant increase of apoptotic cells / crypt loss associated with enhanced lymphocyte and neutrophil infiltration. Whereas the number of CD8 cells in the lamina propria significantly increased after SCT, CD4 cell numbers were strongly diminished. Within post-transplant biopsies, loss of crypts (score 0.5 in GvHD 0-1, 1.2 in GvHD 2 and 1.3 in GvHD 3/4 ), changes in neutrophil infiltrates (score 0.3 in GvHD 0-1, 0.8 in GvHD 2 and 0.1 in GvHD 3/4 ) and reduction of CD4 infiltrates (score 1.20 in GvHD 0-1, 0.85 in GvHD 2 and 0.39 in GvHD 3/4) were clearly correlated with stages of gastrointestinal GvHD, whereas CD8 cells showed an increase and CD25 positive cells were unchanged in pts with severe GvHD. Presence of NOD2/CARD15 SNPs themselves resulted in a significant reduction of neutrophils (p 0.04) and CD4 cells (p 0.006) but had no impact on further parameters. This effect could not be explained indirectly by more severe GvHD in pts with NOD2/CARD15 SNPs and was confirmed in multivariate analysis.

Our data indicate for the first time functional changes in gastrointestinal biopsies from pts after allogeneic SCT in relation to the NOD2/CARD15 genotype. The observed reduction of neutrophils and CD4 cells may result from a reduced expression of chemokines attracting these cells to inflammatory sites, as IL-8 production is strongly regulated by NOD2/CARD15 dependent activation of NF-kappaB. Presence of CD4 cells and neutrophils may be required to prevent dysregulated inflammation.

Introduction: The NIH staging and response criteria offer for the first time criteria for standardized diagnosis and staging of severity as well as evaluation of physical functioning and quality of life (QoL) of chronic graft-versus-host disease (cGVHD). We present the interim analysis of a prospective Germany multicenter validation study on the NIH staging criteria in cGVHD. Methods: 102 patients (median age 45 years, range 18-67) after allogeneic hematopoetic stem cell transplantation (HSCT) for hematologic malignancies were evaluated according to the NIH criteria based cGVHD activity assessment, the Lee cGVHD-symptom-scale, FACT-BMT, human activity profile (HAP), SF 36, Berliner Social Support Scale (BSSS), 24 item Adjective Measure (24-AM), Hospital Anxiety and Depression Scale (HADS) and the NCCN-Distress-Thermometer. Enrolment occurred between day 100 and 1 year after HSCT or in the presence of active cGVHD without time limit. Follow-up surveys were conducted at 1, 2, 3, 5, 8, 12 and 18 months after baseline survey. At all time points disease status, co-morbidities and medication were documented. Results: Sixty five patients had cGVHD (mild n=18, moderate n=26, severe n=21) while 37 patients did not have cGVHD. The comparison of the severity grading (mild-moderatesevere) of the physician and severity grading of the patient revealed a high correlation (p<0.01, r=.66), while the comparison of 10 point scale of patient and physician revealed differences between patient and physician in the range of 3-5 points (physician), where patients graded more severity compared to physicians. The cGVHD NIH consensus grading correlated inversely with FACT physical well being (r=.41, p <0.001). The HAP maximum activity score correlated inversely with severity of cGVHD (p<0.05, r=.35). The cGVHD symptom scale summary score correlated with physician severity grading (r=.66, p<.0001), the FACT-G score (r=.66, p<0.0001), mental health (r=.56), energy and vitality (r=.5) and HAP maximum score (r=.37, p < 0.001). Beside FACT physical and functional well being the HAP maximum score was independent of other aspects of QoL. Discussion: The results demonstrate, that severity of cGVHD as assessed by the NIH consensus grading correlates with impairment of physical well being as well as daily activities and QoL. Since the cGVHD symptom scale covers severity of cGVHD and aspects of QoL it should be applied together with the HAP in clinical routine.

Naturally occurring regulatory T cells (Tregs) have been reported to play an important role in modulating graft-versushost disease (GVHD), a major complication after allogeneic haematopoietic stem cell transplantation. Despite striking findings in animal models supporting the therapeutic use of Tregs in GVHD, the data from human studies is limited and their mechanism of action remains elusive. In this study, Treg modulation of CD8+ lymphocyte induced in situ graft-versushost reactions (GVHR) was evaluated using a unique in vitro human GVHD model. Tregs were defined as CD4+CD25hiFoxp3+ and isolated from buffy coat of healthy blood donors using RoboSep following RosetteSep enrichment of CD4+ cells (Stemcell Technology). Isolated Tregs were expanded in vitro with anti-CD3CD28 mAb coated Dynabeads (Invitrogen) prior to use. The alloreactive immune reactions were set up by co-culturing "donor" CD8+ lymphocytes with HLA unmatched allogeneic "recipient" monocyte derived DCs (mo-DC) in the absence or presence of "donor"-derived Tregs (1:4 ratio for Treg: CD8+ lymphocytes) for 7-8 days. Following magnetic depletion of mo-DC and Tregs, allo-antigen stimulated "donor" CD8+ lymphocytes were co-cultured for 3 days with "recipient" skin tissue. The severity of histopathological changes in skin tissue was scored as grades I-IV according to the Lerner GVHD grading system. In 4 out of 4 experiments the presence of Tregs significantly reduced the severity of skin GVHR from grade III to grade I. The levels of IFNg, TNFa and IL-5 cytokines in the supernatants from the primary co-culture of allogeneic mo-DC and CD8+ lymphocytes were significantly reduced in the presence of Tregs (IFNg: 2179pg/ml vs 40.41pg/ml, p<0.0001; TNFa: 128.2 vs 19.55, p=0.0001 and IL-5: 1084.7 vs 11.52, p=0.0006). Following allogeneic mo-DC stimulation there was a 5.5 and 7.6 fold reduction in the percentage of CD8+ lymphocytes expressing the activation marker CD69+ (13.2 vs 2.4, p=0.026) and intracellular IFNg (18.2 vs 2.5, p=0.029) in the presence of Tregs. CD8+ lymphocyte proliferation measured by 3H-thymidine incorporation and CFSE dilution was found to be markedly suppressed (67%-96% inhibition) in the presence of Tregs in alloreactions. Further investigations are underway to explore the mechanisms and characterise the modulation of GVHR by Tregs in an allo-antigen specific setting.

Donor-recipient HLA class I ligands and KIR-haplotype A are associated with severe acute graft-versus-host disease in unrelated haematopoietic stem cell transplantation for beta-thalassaemia R. Littera (1) Killer immunoglobulin-like receptors (KIRs) regulate the activity of human natural killer cells, mainly through recognition of HLA Class I molecules. Two broad haplotypes of KIR genes have been defined. The A haplotype is characterised by a single activating KIR gene (2DS4), whereas the B haplotype is characterised by two or more activating KIR genes (2DS1, 2DS2, 2DS3, 2DS5 and 3DS1). Many studies have investigated the impact of KIRs and their ligands on hematopoietic stem cell transplantation (HSCT) in patients affected by acute myeloid or acute lymphoblastic leukemia. However, the results of these studies remain controversial. Allogeneic HSCT in talassemia patients offers an ideal study model since this cohort of patients is not biased by the variability of conditioning regimens and the different clinical and immunologic characteristics of patients transplanted for oncohematologic disorders. We studied 66 thalassemia patients transplanted from an unrelated donor. The conditioning regimen was the same in all patients. Donor and recipient pairs were typed for the HLA-A, B, Cw, DRB1, DRB3, DRB4, DRB5, DQA1, DQB1 and DPB1 loci using high resolution molecular typing techniques. KIR genes were typed using KIR-gene-specific primers. Out of 66 transplanted patients, 52 are alive and well (disease-free survival 78.7%), 8 rejected and 6 died. Twentytwo patients (22/66 -33.3%) developed acute graft-versushost disease (aGvHD). In 6 of these patients aGvHD was Grade III-IV. Patients who were heterozygous for HLA-Cw groups 1 (HLA-CwAsn80) and 2 (HLA-CwLys80) had a higher risk of developing acute GvHD than C1/C1 or C2/C2 homozygotes (14/29 vs 8/37; RR=3.3; 95% CI: 1.63-9.76; p=.03). Conversely, 7/8 patients who rejected were C1/C1 or C2/C2 homozygotes (RR=6.5; 95% CI = 0.75 -56.54; p=.06 when compared with heterozygotes). These findings confirm the results of a previous study performed on a cohort of 45 thalassemia patients. In the present study, 5 of the 6 patients (83.3%) with severe Grade III-IV aGvHD had been transplanted from donors who were homozygous for KIR haplotype A (RR=23.4; 95% CI: 1.19-457.96; p=.008). In conclusion, it would seem that C1/C2 heterozygosity associated with donor homozygosity for the A haplotype is likely to favour donor alloreactivity and thereby increase the risk of severe GvHD. Analyses of these genetic markers may help modulate conditioning regimens and the intensity of GvHD prophylaxis in patients undergoing unrelated HSCT.

Interleukin-13 (IL-13) is an immunoregulatory cytokine secreted predominantly by activated T-helper 2 (TH2) cells. It suppresses the cytotoxic action of macrophages, inhibits the production of pro-inflammatory cytokines and is a central mediator of allergic inflammation. A single nucleotide polymorphism (SNP) exists within exon 4 of the IL13 gene at position +2044. The A allele of this SNP and high IL-13 levels have been linked with several inflammatory conditions and IL-13 mixed lymphocyte culture (MLC) levels have been associated with graft-versus-host disease (GVHD) following haematopoietic stem cell transplantation (HSCT). Consequently the roles of the IL13 +2044 SNP and IL-13 MLC levels in HSCT were examined in this investigation. Polymorphism studies were carried out on a cohort of 923 HSCT recipients and donors from 7 transplant centres across Europe. IL13 genotyping was performed using PCR and RFLP analysis. IL-13 levels were measured in a cohort of 91 MLC supernatants using a cytometric bead array and correlated with GVH reaction (GVHR) grades from an in vitro model of GVHD. In all statistical analyses P values <0.05 were regarded as being significant. Multivariant analysis of the whole HSCT cohort demonstrated that the IL13 +2044 A allele was significantly associated with the development of both acute (grades III-IV) and chronic GVHD (P=0.028 and P=0.026 respectively). These associations remained significant when the cohort was stratified for transplant type and conditioning regimen. Significant associations were also observed in a subset of patients diagnosed with CML; in HLA-matched siblings possession of the IL13 +2044 A allele was linked with a decreased susceptibility to chronic GVHD, whereas in MUD transplants possession of the A allele was a risk for chronic GVHD. Depending on the subset analysis, several clinical factors were also significantly associated with GVHD. Analysis of the MLC data demonstrated that IL-13 levels increased with GVHR grade, with significantly higher levels being observed in MLC supernatants with GVHR grades III-IV (P=0.015). To our knowledge this is the first investigation examining the roles of both IL-13 MLC levels and the +2044 SNP in HSCT. The findings are encouraging, indicating that IL-13 may be involved in the immunopathology of GVHD. Consequently, IL-13 levels, as well as SNP analysis could provide key pretransplant information on GVHD prognosis and be potential novel targets for post-HSCT GVHD therapy. 

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is a curative treatment for hematologic malignancies, but the application is slimited due to major complications, such as severe graft versus host disease (GvHD). Diagnosis of acute GvHD is based on clinical features and biopsies, a proteomic pattern specific for aGvHD has been published and evaluated blindly on 902 samples collected from 168 patients undergoing allo-HSCT at MHH between 2005 and 2007. The majority of the patients included were transplanted for hematological malignancies (n=158), 10 for hematopoietic failure syndromes (sAA, n=6; PNH n=1; OMF, n=3). Forty-five patients were treated with dose-reduced conditioning regimens; GvHD-prophylaxis consisted of cyclosporin (CsA) plus methotrexate (MTX) or mycophenolic acid (MMF), and antibodies respectively. Most patients were transplanted from matched unrelated donors (MUD, n=100),63 received stem cells from matched related (MRD, SIB, 1 syngeneic), 3 from haplo-identical related, and 2 from mismatched related donors. Based on the positivity of the aGvHD pattern we initiated a pilot trial of pre-emptive therapy, treating patients upon pattern positivity with 1mg prednisone/kg BW and compared the outcome of the treated group to those patients who did not receive pre-emptive therapy. In 2005 90 patients were transplanted and screened for aGvHD at MHH, but not pre-emptively treated. Forty one patients developed aGvHD (45%), 22 (24%) had aGvHD >II and were treated with standard protocols. Eight patients developed aGvHD grade III or IV (8/22=36%) and 6 of these died. Between April 2006 and June 2007 78 patients were transplanted at MHH. In 30 patients the aGvHD proteomic pattern showed a clear correlation for GvHD >II. Twenty six (33%) of 78 transplanted patients had aGvHD >II, 12 received pre-emptive therapy, while 14 were treated upon clinical signs of aGvHD according to standard treatment protocols.Two patients in the preemptive treatment group (2/26=7.9%; 2 of 12 pre-emptively treated:16 %) had aGvHD III or IV and died (2/26=7.9%; 2/12: 16%). Of the 14 standard therapy patients 6 developed aGvHD grade III or IV and to date 3 of these have died (6/26 =23% or 6/14 of the standard therapy group=42%). Thus, taken together our results indicate that pre-emptive treatment may decrease the severity of aGvHD and probably leads to a better overall survival.

Donor-derived T cells emigrating from the graft after solid organ transplantation have been shown to promote immunological tolerance thereby improving long-term graft survival. However, donor T cells are also able to induce lifethreatening allo-reactive graft-versus-host disease (GVHD) in the transplant recipient. The exact mechanism by which donor T cells influences this delicate balance between tolerogenicity and allo-reactivity has not been elucidated. We observed two liver transplant patients with severe GVHD who developed a donor T-cell chimerism in peripheral blood and bone marrow up to 100%, which is far above those of previously described cases. This enabled us to isolate donorderived T cells in sufficient numbers to allow for a detailed analysis of phenotypic and functional features ex vivo. We found that the donor T cells died by apoptosis over time without any evidence of rejection by host T cells. Interestingly, the host-versus-donor reactivity appeared to be selectively impaired, as anti-viral T cells were still detectable in the host repertoire. These results indicate that graft T cells are able to specifically eliminate donor-reactive T cells from the host repertoire thereby preventing donor cells from subsequent rejection. Since substantial donor T-cell chimerism persisted in both patients beyond resolution of GVHD, we investigated potential mechanisms of immunotolerance. We observed that the recovery from GVHD was not accompanied by an expansion of immunosuppressive CD4/CD25/FoxP3-positive T cells in peripheral blood. However, we obtained formal evidence that host-reactive donor T cells were controlled by an alternative negative regulatory pathway, executed by the immunoinhibitory receptor programmed death-1 (PD-1) and its ligand PD-L1. We did not only find an exceptionally high level of PD-1 expression on host-reactive donor T cells ex vivo, but also discovered that blocking PD-L1 on host cells significantly enhanced anti-host reactivity by donor CD8 T cells in vitro. We thus suggest the interference with the PD1/PD-L1 pathway as a novel therapeutic strategy to control host-reactive donor T cells in solid organ transplant-associated GVHD. Our observations might be also of relevance for other clinical scenarios of misdirected allo-reactivity, such as graft rejection as well as severe GVHD after allogeneic hematopoietic stemcell transplantation. (2) Purpose: To determine risk factors of outcomes after umbilical cord blood transplantation (UCBT) for patients with advanced lymphoid malignancies. Patients and methods: We evaluated 104 adult patients (median age, 41 years) who underwent unrelated donor UCBT for lymphoid malignancies. UCB grafts were 2 antigen HLA mismatched in 61%, and were composed of one (n=78) or two (n=26) units, with a median cell dose of 2.5x10 7 nucleated cells/kg and 1.05x10 5 CD34 cells/kg. Diagnoses were non-Hodgkin lymphoma (NHL, n=62), Hodgkin lymphoma (HL, n=29), and chronic lymphocytic leukemia (CLL, n=13), with 85% having advanced disease and 60% having failed a prior autologous transplant. Sixty-four percent of patients received a reduced-intensity conditioning regimen and 44% low-dose total body irradiation (TBI). Median follow-up was 14 months. Results: Cumulative incidence (CI) of neutrophil engraftment was 85% by day 60, with greater engraftment in recipients of higher CD34+/kg cell dose (93% vs. 78%, p=0.0001). CI of non-relapse-related mortality (NRM) was 28% at 1 year, with a lower risk in patients treated with low-dose TBI (13% vs. 47%, p=0.007). CI of relapse or progression was 31% at 1 year, with a lower risk in recipients of double unit UCBT (9% vs. 38%, p=0.02), and those with chemosensitive disease (19% vs. 40%, p=0 .01) and indolent NHL (19% vs. 35%, p=0.04) . The probability of progression-free survival (PFS) was 41% at 1 year, with higher survival in those with indolent NHL (61% vs. 34%, p=0.04), chemosensitive disease (55% vs. 31%, p=0.004) and who received low-dose TBI (58% vs. 27%, p=0.002). Conclusion: UCBT is a viable treatment for adults with advanced lymphoid malignancies. Diagnosis of indolent lymphoma, chemosensitive diseases, and use of low-dose TBI and were factors associated with significantly better outcome.

Positron emission tomography scan performed before reduced-intensity conditioning allogeneic stem cell transplantation has a prognostic value in patients with relapsed and chemosensitive Hodgkin's lymphoma or aggressive non-Hodgkin lymphoma A. Dodero* (1), R. Crocchiolo (2) (1) (

Positron emission tomography (PET) scan using 18fluorodeoxyglucose [18F-FDG] has a recognised prognostic value in patients (pts) with Hodgkin Lymphoma (HL) or aggressive Non-Hodgkin Lymphoma (HG-NHL) receiving chemotherapy or autologous stem cell transplantation (SCT). Thus, we retrospectively assessed the prognostic role of PET scan before reduced-intensity conditioning allogeneic SCT. Between 2000 and 2007, 82 consecutive pts with HG-NHL or HL, responding to salvage therapy, were evaluated with a PET scan before allografting. Presence (PET-pos) or absence (PET-neg) of abnormal 18F-FDG uptake was correlated to progression-free survival (PFS) and overall survival (OS).Interpretation of PET scan was obtained with visual assessment alone by a nuclear medicine physician (evaluation of maximal SUV in PET-pos cases is ongoing). Median age of pts was 36 years (range, 17 -68 years). Histologic subtypes included: 38 HG-NHL [B phenotype (n=25), T phenotype (n=12), other (n=1)] and 44 HL. Fortyseven pts (57%) were allografted from a HLA-identical sibling donor, 16 from a haploidentical donor and 19 from an unrelated donor. Sixty-eight pts (83%) failed autograft, the median number of prior regimens was 3 (range, 1-6). PET scans were performed at a median of 30 days prior to allograft: 41 out of 82 pts were PET-pos [HG-NHL (n=18), HL (n=23)] whereas 41 were PET-neg [HG-NHL (n=20), HL (n=21)]. Pts with PET-pos or PET-neg scans were well balanced in terms of diagnosis, previous treatments, and type of donor. At a median follow-up of 30 months (range, 6 -86 months), 54 pts are alive and 28 died [toxicity n=12, disease n=16]. Overall, the estimated 3-year PFS in pts with PET-neg or PET-pos scans were 68% (95% CI, 49% -81%) versus 30% (95% CI, 15% -47%), respectively (p<0.003). For HG-NHL pts, the estimated 3-year PFS was 70% for PET-neg as compared to 41% for PET-pos (p<0.02) whereas for HL pts, the estimated 3-year PFS was 68% as compared to 17%, respectively (p=0.05). A statistically significant higher cumulative risk of relapse was observed in pts with PET-pos scan before allograft as compared to the PET neg scan (53% versus 21%, p< 0.022). The estimated 3-year OS in pts with neg or pos PET scans were 77% (95% CI; 60% -87%) versus 41% (95% CI; 24%-57%), respectively (p< 0.002). Our study shows a better PFS and OS for pts being PET neg before allografting. PET scan should be incorporated in pretransplant work-up to validate our findings prospectively. L. Rigacci*, A. Bosi, B. Puccini, P. Corradini, L. Castagna, N. Cascavilla, G. Milone, A. Bacigalupo, R. Scimè, G. Specchia, A. Rambaldi, P. Leoni, F. Ciceri, A. Levis, S. Guidi, B. Bruno, R 

Diffuse large B cell lymphoma (DLBCL) is the most common lymphoid malignancy in adults. Autologous hematopoietic stem cell transplantation (AHSCT) has been shown to be an effective therapy for patients with DLBCL who relapsed after complete remission (CR). Patients who relapse after an AHSCT have a very poor prognosis and usually can not be cured with standard or high dose chemotherapy. Allogeneic hematopoietic stem cell transplantation (AlloHSCT) has shown to be effective in the rescue of lymphoma patients relapsed after conventional or high dose therapies. According to this data we have analysed all patients with diagnosis of DLBCL who have performed an AlloHSCT from 2000 to 2005 after an AHSCT relapse. Seventy-five patients were selected from the data-base, 46 were male (61%), 71 presented a diagnosis of DLBCL and 4 were Anaplastic large cell lymphoma. The stem cell donor was related in 61 patients (81%) and unrelated in 14 patients (19%). The stem cell source was peripheral blood in 61 cases and bone marrow in 14 cases. The conditioning regimen was conventional in 24 patients and reduced intensity in 51 patients. The median time between AHSCT and AlloHSCT was 13 months (range 1-68 months). Twenty-five patients (33%) performed AlloHSCT after the obtainment of at least a partial remission or a controlled disease, 36 (48%) were treated with active disease and in 14 cases the data was not available. After AlloHSCT 75% of patients obtained a response (CR or PR) and did not have evidence of disease, 25% of patients did not respond and progressed. The treatment related mortality (TRM) was 32%, 14 out 24 (58%) patients died in conventional regimen and 10 out 51 (20%) in reduced intensity arm. Acute graft versus host disease (aGVHD) was observed in 28 patients (grade III-IV in 8 patients), and chronic (cGVHD) in 65%. After a median follow-up of 58 months from the diagnosis (range 11-196 months) and a median follow-up of 9 months after AlloHSCT (range 0-82 months) the overall survival was 45%. The overall survival was significantly higher in patients treated with reduced intensity conditioning in comparison with patients treated with conventional conditioning (p:0.001). This retrospective study confirms that allo-transplant it is feasible and it could be really effective in a poor prognosis group of patients. Moreover the use of reduced intensity conditioning improves these results. (2) Median time from diagnosis to MUD-SCT was 25 months (range, 3 -205). 64% of the patients had failed previous autologous transplant (ASCT), and 25% were transplanted with chemorefractory disease. Peripheral blood was the source of hematopoietic stem cells in 70% and reduced intensity conditioning regimens (RIC) were used in 52% of the cases. After a median follow up for living patients of 35 months, the estimated 3-year non-relapse mortality (NRM), relapse rate (RR), progression free survival (PFS) and overall survival (OS) for the whole series were 32%, 38%, 30% and 38.5%, respectively. Grade II-IV acute graft-versus-hostdisease developed in 32% of patients. Patients selected for RIC protocols were older (median age of 44 years vs 38 years, p = 0.02) and more heavily pre-treated; 75% had failed autograft compared with 53% in the conventional conditioning(CC) group (p = 0.01). Despite these unfavorable factors, NRM for patients receiving RIC was significantly lower than observed in patients treated with a conventional regimen:23% vs 41% at 3 years (p = 0.02). However, this advantage was offset by an increased RR in patients undergoing RIC-MUD (3-yr RR: 46% vs 30%, p = 0.2), resulting in a very similar PFS and OS for both types of conditioning regimens. The prognostic factor with highest impact on PFS was refractory disease at transplantation (RR = 1.8; 95%CI 1.1 -3.1, p = 0.02). Patients transplanted with chemosensitive disease had a 3 yr PFS of 35% irrespective to the conditioning regimen applied, whereas patients transplanted with chemorefractory disease had a 3 year PFS of 16% only. In conclusion, MUD SCT provides a true chance for cure for select patients with DLBC lymphoma who failed conventional therapies, particularity if transplanted with cemosensitive disease. RIC, knowm to be associated with a reduced NRM rate, should be especially considered in patients with chemosensitive patients, whereas CC might be the preferred option for patients with more aggressive disease.

Prospective evaluation of 18F-fluorodeoxyglucose (FDG) positron emission tomography as a predictor of residual disease and subsequent relapse in patients with diffuse large cell ymphoma and Hodgkin's lymphoma undergoing HDC and ASCT S. Akhtar*, A. Al-Sugair, Y. Al Kadhi, A. Al-Zahrani, M. Abdelsalam, S. Bazarbashi, D. Ajarim, I. Maghfoor King Faisal Specialist Hosp. & Res. Centre (Riyadh, SA) Background: There is emerging data indicating poor outcome in diffuse large cell ymphoma (DLCL) and Hodgkins Lymphoma (HL) patients with positive 18F-fluorodeoxyglucose (FDG) positron emission tomography (PET) before high dose chemotherapy (HDC) and autologous stem cell transplant (ASCT). We initiated this prospective trial to evaluate the impact of PET as a predictor of post HDC residual disease and relapse in patients with DLCL and HL undergoing HDC ASCT. Patients and methods: From July 2005 to June 2007, 115 patients with relapsed or refractory DLCL and HL were enrolled as a potential candidate for HDC ASCT. 43 patients did not have HDC ASCT due to progression (10), refusal (4), noncompliance (5) or other reasons (24). All eligible patients received ESHAP as salvage chemotherapy, responding DLCL or responding / stable HL patients received same chemotherapy as mobilization regimen for stem cell collection followed by BEAM as HDC. Prior to the initiation of ESHAP, each patient had CT scan + other radiological studies if needed and PET ( CT-1 and PET-1), after 2-3 cycles of salvage chemotherapy / prior to HDC ASCT (CT-2 and PET-2) and approximately 100 days post HDC ASCT or earlier if clinically indicated (CT-3 and PET-3). 72 patients had HDC ASCT and 49 of them had both PET-2 and PET-3 available at the time of this analysis. Results: Patients characteristics are male: 28, female: 21, DLCL: 9 and HL: 37. Relapsed 21 and refractory 28. Median age at ASCT is 32 years (16 to 62). Median follow-up from ASCT is 12 months. 27 patients were PET-2 negative prior to HDC ASCT. At the time of this evaluation, of these 27 patients, 4 (15 %) had an event, event free survival (EFS) 85%. 22 patients were PET-2 positive prior to HDC ASCT, 10 (46 %) had an event, EFS 54%. EFS for PET-2 negative vs PET-2 positive has P value of 0.027. Using Kaplan-Meier method, positive PET-2 has 53% probability of an event vs 16% for a negative PET scan (P=0.005). EFS for CT-2 negative vs CT-2 positive patients is 83% vs 68%, P = 0.466. Kaplan-Meier for CT-2 positive showed 36 % probability of event vs 18% for CT-2 negative, P=0.198. Conclusions: Prior to HDC ASCT, positive PET scan indicates high risk of residual disease or progression. Many of these patients are likely to suffer from treatment failure. These patients are potential candidate for more aggressive and experimental therapies. For many patient with pre-transplant negative PET, post transplant PET scan can be omitted.

Radioimmunotherapy with haemopoietic stem cell transplantation for treatment of malignant non-Hodgkin lymphoma: multicentre study on thirty patients A. Mele (1) (7), G. Console (7), N. Cascavilla (2) , P. Scalzulli (2) /Kilograms (range 2.55-21.6). All patients engrafted. The median number of red blood cell and platelet transfusion were 4 (1-7) and 6 (1-8), respectively. The median time to platelet and neutrophil counts higher than 20x10 9 /L and 0.5x10 9 /L were 14 (range, 9-35 days) and 10 days (range, 8-14) , respectively. Mucosites occurred in all patients (grade III and IV in 13 and 4 cases). Febrile neutropenia occurred in 80% of cases. Six pneumonitis and 8 blood stream infections were documented. One patient developed an atrial fibrillation. Twenty-one of 30 patients were valuable for 90-day response. The 90-day overall response was 86% with 72% of CR. Four early deaths before day-90 occurred: 1 case for septic shock (day +6), 1 for viral encephalites (day +60) and 2 for progression disease (day + 30, 65). The Kaplan-Meyer estimated treatment related mortality (TRM) is 9%. Seven cases (1 CR, 4 PR and 2 not response) progressed at a median follow-up of 95 days post HST (range, 60-300). Twenty-four of 30 patients are alive at a median follow-up of 175 days post HST (range, 6-590). Six patients died (20%): 3 for progression, 1 in CR for ARDS (day + 230) and 2 for TRM. Sixteen (53%), 5 (17%) and 2 of 30 patients are alive in CR, PR and progression, respectively. One case is not valuable for response (day+15). We analyzed the characteristics of 14 alive patients in CR: 9 had aggressive lymphoma and 13 were at least in PR before HST (p=NS). The Kaplan-Meyer estimated 2y-EFS is 73%. Conclusion. The use of RIT plus transplant induces 70% of OR (53% CR) with sustained engraftment, an acceptable extra-haematological toxicity and a rapid immunological recovery in patients who failed to achieve CR after first line chemotherapy. The power of this program needs to be assessed in a larger series of patients. HDT/ASCT has been planned either in front-line pts with high prognostic score/Bulky mass/stage III-IV or at relapse. First line therapy were mainly ABVD/MINE for HD and ACVBP/CHOP ± Rituximab for NHL. FDG-PET was performed after 1-5 chemotherapy cure. MINE and DHAP were the most frequent salvage chemotherapy used for refractory front-line therapy pts. The conditioning regimen was mainly BICNU-etoposide-aracytine-melphalan. 46 pts (53%) were pre-HDT/ASCT PET negative and 41 positive (47%). 8/41 pre-HDT/ASCT positive pts (19.5%) converted to negative by additional cross chemotherapy. After HDT/ASCT, 22/33 others pre-HDT/ASCT positive pts (67%) converted to post-HDT/ASCT PET negative. One negative pre-HDT/ASCT PET converted to positive. Residual disease of positive pts was mainly treated by local radiation. After a median follow-up of 3.2 years (range 0.4-8.4) after pre-HDT/ASCT PET, 29 pts relapsed, 21 dead (with 18 of 29 relapses), 3 remained resistant disease. Survival was measured from pre-HDT/ASCT PET to death (overall survival OS) or relapse/death (event-free survival EFS) with censoring time at the time of last follow-up. Median OS and EFS for the two groups were not reached. Estimated 3 years OS was 80% and 73%, respectively for pre-HDT/ASCT PET negative and positive pts. Estimated 3 years EFS was 74% and 61% respectively for pre-HDT/ASCT PET negative and positive groups.

A positive FDG-PET after induction chemotherapy is highly predictive of poor survival in HD and NHL pts but an additional risk-adapted treatment strategies before HDT/ASCT by salvage cross chemotherapy and after HDT/ASCT by targeted radiation may improve pre-HDT/ASCT PET positive pts outcome.

Donor lymphocyte transfusions (DLT) after allogeneic stem cell transplantation have been shown to be very effective in treatment of recurrent mylogenous leukemia but displayed limited use in chronic lymphocytic leukemia (CLL) and highly malignant non-Hodgkin lymphoma (NHL). Here we studied whether Bi20 (FBTA05), a novel trifunctional bispecific antibody targeting CD20 on lymphoma cells and CD3 on T cells could induce graft-versus-leukemia / lymphoma responses in combination with DLT or mobilized peripheral blood stem cells (PBSCT) after allogeneic transplantation in these diseases. Six patients (3 cases each with p53 mutated CLL, and highgrade NHL) refractory to standard therapy were treated with escalating doses of Bi20 (range 10 -2000 µg) followed by DLT or SCT. In 4 out of 6 patients, a prompt, but transient clinical and hematological response was observed. Side effects (fever, chills and bone pain) were tolerable and appeared at lower dose levels in CLL (>40 µg) than in high grade NHL (>200 µg). The cytokine profile was characterized by transient increases of IL-6, IL-8 and IL-10. Neither human anti-mouse antibodies (HAMAs) nor graft-versus-host disease (GVHD) developed allowing repeated treatment courses. In summary, the trifunctional antibody Bi20 induced prompt antitumor responses in extensively pretreated, p53 mutated, alemtuzumab and rituximab refractory patients indicating its therapeutic potential. D. Stachel*, K. Kirby, L. Corey, M. Boeckh Fred Hutchinson Cancer Research Center (Seattle, US) Background: Although CMV viral load is a predictor of CMV disease in HCT recipients, regulatory agencies presently do not accept it as primary endpoint for studies that evaluate new therapeutics. The aim of this study was to examine whether CMV viral load predicts transplant-related mortality (TRM) or overall survival (OS) in the era of preemptive therapy. Methods: 2896 consecutive patients following a first HCT at FHCRC between 1995 and 2005 were analyzed; 1481 of them were CMV seropositive. Patients underwent weekly testing for CMV viremia by culture and quantitative pp65 antigenemia (AG) during the first 100 days after HCT. Preemptive antiviral therapy was given for any pp 65 AG. Using univariate and multivariable COX models, we analyzed the association of initial, mean, peak CMV load, and CMV area under the curve (AUC) for their association with OS and TRM at 1 year after HCT. Viral load parameters were analyzed in quartiles. Pp65 AG results are expressed as cells/200,000 cells. Results: TRM occurred in 25.2% of all patients and 27.8% of seropositive recipients.

In a model that included CMV seropositive recipients, the adjusted hazard ratios (HR) for the upper quartile of initial viral load (AG > 5), peak viral load (AG >10) and AUC (AG>100) were 1.6 (95% CI 1.2-2.1), 2.2 (95% CI 1.6-3.0), and 1.7 (95% CI 1.3-2.3), respectively. The statistical models were adjusted for recipient and donor age, HLA match, donor relationship, year of HCT, acute/chronic GvHD and postengraftment neutropenia (to account for the toxicity associated with preemptive therapy); additional factors evaluated (but not significant in univariate analysis) were race, donor CMV serostatus, cell source, type and intensity of conditioning, T cell depletion, risk of the underlying disease, and GvHD prophylaxis. Similar models that included all patients and those that used OS as endpoint also showed significant associations for first and peak viral load and the AUC. Results are presently being validated in a separate cohort that underwent PCR surveillance. Conclusion: Initial and peak viral load as well as the viral AUC are independently associated with TRM and OS in allogeneic HCT recipients receiving preemptive antiviral therapy. The peak viral load showed the strongest association with TRM and OS. These data further support the use of parameters of viral dynamics as primary endpoints for studies that evaluate immune augmentation or drug prevention strategies.

Objectives: To evaluate the effect of immunoglobulin (IVIG) and cytomegalovirus-hyperimmune immunoglobulin (CMV-IVIG) prophylaxis in patients undergoing hematopoietic stem cell transplantation (HSCT). Methods: Systematic review and meta-analysis of randomized controlled trials comparing systemic IVIG or CMV-IVIG with placebo or no intervention for prophylaxis in patients undergoing HSCT. The Cochrane Library, MEDLINE, conference proceedings and references were searched until 2007. The primary outcome was all-cause-mortality at end of follow-up. Two reviewers independently appraised the quality of the trials and extracted data. Relative risks (RR) with 95% confidence intervals were estimated and pooled. Results: Twenty trials fulfilled inclusion criteria, 12 assessed IVIG and 8 assessed CMV-IVIG. For IVIG vs. control there was no difference in all-cause mortality, RR 0.99; 0.88-1.12, 8 trials, Fig 1. Results were similar when only patients following allogeneic HSCT were assessed (3 trials), when mortality was assessed at day 100-200 (5 trials) or after 2 years (3 trials), and when high-dose IVIG was used (3 trials). Trials' methodological quality did not impact results. Overall, there was a reduction in the number episodes of interstitial pneumonitis (IP), RR 0.64; 0.45-0.89, 7 trials, although the reduction in CMV infections was not statistically significant, RR 0.84; 0.66-1.07, 6 trials. There was no difference in clinically or microbiologically documented infections, RR 1.00; 0.90-1.10 and RR 1.00; 0.88-1.15, respectively, 7 trials both). There was no difference with regard to acute graft vs. host disease (GVHD), RR 0.93; 0.83-1.04, 7 trials. Veno-occlusive disease (VOD) was significantly more frequent with IVIG, RR 2.73; 1.11-6.71, 4 trials, Fig 2, 

An open-label randomised study of oral valganciclovir versus intravenous ganciclovir for pre-emptive therapy of cytomegalovirus infection after allogeneic stem cell transplantation L. Volin* (1), L. Barkholt (2) 

Cytomegalovirus (CMV) remains a leading cause of infectious complications after allogeneic stem cell transplantation (SCT) and without preventive measures patients with CMV infection have a significant risk to develope CMV disease. A preemptive treatment approach with iv-ganciclovir or ivfoscarnet is used to restrict the use of potentially toxic drugs for patients at risk for CMV disease. Recently, valgan-ciclovir, an oral prodrug of ganciclovir, has become available. The Nordic BMT group started a study with the aim of demonstrating that treatment outcome of the first CMV DNAemia after SCT with oral valganciclovir is not inferior to that obtained with iv-ganciclovir. The primary study endpoint was achievement of quantitative CMV PCR (qPCR) negativity on day 28 of treatment or earlier. qPCR was carried out weekly for the first three months after SCT, and if positive, patients were randomized to receive either oral valganciclovir 900 mg twice a day or ganciclovir iv 5 mg/kg every 12 h for 14 days (adjusted according to renal function). qPCR was studied twice a week. If qPCR was negative on day 13-14 of treatment, the treatment was discontinued, and if negativity was not achieved but the copy number was decreasing, maintenance treatment once daily was administered for another 14 days. Thereafter the treatment was carried out according to local policy. Of the 96 patients receiving an allogeneic SCT, 74 patients were treated for a hematological malignancy and 22 for a solid tumor. The patients in the poand iv-groups did not differ by recipient/donor CMV serostatus, age (median 55 vs 51 years), diagnosis, donor (sibling/ unrelated), or the time of PCR positivity (day +40 and +38 after SCT, respectively). The median CMV DNA copy number/ml at diagnosis was 2525 (200-160000) in the pogroup, and 1425 (200-41900) in the iv-group. The maximum copy number/ml was 5031 (300-160000), and 2750 (200-41900), respectively. The incidence of aGVHD ≥ grade 2 was 30% in the po-group and 19% in the iv-group. The primary endpoint of the study was achieved in 30/49 (61%) of the popatients and 25/47 (53%) of the iv-patients. During the study three patients with a solid tumor de-veloped CMV disease, two in the po-group and one in the iv-group. Oral valganciclovir therapy seems not to be inferior to iv-ganciclovir therapy in the preemptive treatment of the first CMV DNAemia after allogeneic SCT, and the practical advantages of oral treatment are obvious.

Genetic variations in the NOD2/CARD15 gene are associated with bacteraemia and sepsis after allogeneic stem cell transplantation M. Grube, D. Veith, G. Rogler, J. Brenmoehl, H. Bremm, J. Hahn, R. Andreesen, E. Holler University Hospital Regensburg (Regensburg, DE) Purpose: Single nucleotide polymorphisms (SNPs) of the NOD/CARD15 gene resulting in a diminished nuclear factor-kappaB (NF-kappaB) response to bacterial cell wall products have been recently associated with an increased incidence of Crohn`s disease as well as transplant related mortality and GvHD following allogeneic transplantation. In addition, a recent study from our group revealed NOD2/CARD15 variants as independent predictors of death from septic shock in nonhematological patients. Therefore, we now analysed the direct interaction of NOD2/CARD15 variants with febrile bacteremia and sepsis in patients receiving allogeneic SCT. Experimental Design: We retrospectively analyzed 139 donor/recipient pairs for single nucleotide polymorphisms (SNPs) of the NOD/CARD15 gene (SNP8, 12 and 13). Bacteremia was determined by bacterial growth in the blood culture. Septic syndrome was determined by classical criteria. Results: 94/139 (68%) patients had unmutated SNP`s (wild type group), in 45/139 (32%) patients either the recipient or the donor had any mutated SNP (R or D any mutation). 68/139 (49%) patients showed febrile bacteremia, whereas 17/68 (25%) developed lethal sepsis syndrome. The cumulative incidence of bacteremia was 48% in the wilde type group and 59% in patients with any mutation (R or D any mutation). The cumulative incidence of lethal sepsis syndrome was 5% in the wilde type group and 39% in the patients with any mutation (R or D any mutation) (p< 0,001). In a more detailed analysis of individual donor and recipient SNPs, presence of SNP8 in the recipient and any SNP in the donor (either SNP8, 12 or 13) were signifikant risk factors whereas the highest cumulative incidence of sepsis was found if the donor had a SNP13 mutation (p < 0,001). Since bacteremia/sepsis syndrome even occurred in patients without GvHD or prior to onset of GvHD our data argue against an induction of bacteremia/sepsis syndrome secondary to GvHD. In line with this view, a multivariate analysis including GvHD showed SNP`s of the NOD/CARD15 gene as independent risk factors for lethal sepsis syndrome (p<0.01). Conclusion: Our results suggest that defective signalling of NOD2/CARD15 either in epithelia (recipient) or monocytes/macrophages (donor) may be directly involved in a diminished antibacterial defense. Translocation of bacteria may be an important step in subsequent SCT related complications.

Infection control interventions for cancer patients: efficacy evaluation through systematic review and metaanalysis M. Paul*, A. Schlesinger, A. Gafter-Gvili, L. Leibovici Rabin Medical Center; Beilinson Campus (Petah-Tikva, IL) Background: Currently used infection control measures for hematological cancer patients are frequently applied with an unclear evidence basis resulting in a heterogeneous application of these interventions. Methods: Systematic review and meta-analysis. Included were all prospective comparative studies assessing the effects of non-pharmacological interventions applied for infection prevention and control among cancer patients following chemotherapy. Interventions were classified according to the transmission modality targeted by the intervention; airborne, contact or endogenous flora. The primary outcome assessed for all interventions was 100-day all-cause mortality. Relative risks (RR) with 95% confidence intervals are reported. Results: We identified 23 studies assessing protective environment (PE) including high particulate air filtration (HEPA) ± laminar airflow (13 randomized trials, 12 hematological cancer, 9 for HSCT), 10 studies comparing outpatient vs. inpatient management for HSCT (all nonrandomized) and 5 miscellaneous studies assessing diet, footwear, gowns or specific room design (3 randomized trials, 2 for HSCT). PE resulted in a remarkable reduction in allcause mortality at day 100: RR 0.79, 0.73-0.87 overall and RR 0.78, 0.66-0.92 in adequately randomized trials alone. The RR was 0.60, 0.50-0.72 for 30-day mortality and 0.86, 0.81-0.91 for the longest follow-up, up to 5 years. Similar survival benefit was observed for allogeneic HSCT patients and other patients. Significant reductions were observed for all infection-related outcomes, except mold infections (Table) . When isolating the effects of the different transmission modalities it became apparent that endogenous suppression using antibacterial (usually combined with antifungal) prophylaxis was the major contributor to the beneficial effect of PE ( Figure) . Outpatient HSCT (3 allogeneic and 7 autologous) was non-inferior to inpatient treatment, RR 0.81, 0.46-1.45 for mortality; RR 0.93, 0.67-1.29 for infections; and RR 0.35, 0.15-0.80 for bacteremia. Conclusions: Protective isolation offers an overall significant benefit to the patient, since the assessment of all-cause mortality encompasses both infection, cancer and treatmentrelated outcomes. The additional value of HEPA and strict contact isolation over antibiotic prophylaxis is unclear and probably depends on the local prevalence of mold infections. Outpatient HSCT appears safe and should be explored in randomized trials.

M. Dettenkofer*, R. Babikir, H. Bertz, A.F. Widmer, W.V. Kern, E. Meyer, P 

For surveillance of nosocomial bloodstream infections (BSI) and pneumonia during neutropenia in adult patients undergoing bone marrow transplantation (BMT) or peripheral blood-stem cell transplantation (PBSCT), an ongoing multicenter surveillance project was initiated by the German National Reference Centre for Surveillance of Nosocomial Infections in 2000 (ONKO-KISS). Methods: Nosocomial Infections are identified using CDC definitions for laboratory-confirmed BSI and modified criteria for pneumonia in neutropenic patients [for detailed information see : CID 2005; 40: 926-31, or in German language: http://www.nrz-hygiene.de/surveillance/onko.htm]. Results: Over the 60-month period from July 2002 up to June 2007 26 centres participated. Altogether 4,909 patients with 72,449 neutropenic days were investigated. Of these, 2,873 (59%) had undergone allogeneic and 2,036 (41%) autologous BMT or PBSCT. The mean length of neutropenia was 14.8 days (9.2 d after autologous and 18.7 d after allogeneic transplantation). In total, 827 bloodstream infections and 403 cases of pneumonia were identified. Site-specific incidence densities are shown in the Table. There was a trend to lower incidence densities over the five years reported. Following allogeneic transplantation, 17.7 BSI/100 patients and 10.4 cases of pneumonia/100 patients occurred whereas following autologous transplantation 15.7 cases of BSI/100 patients and 5.1 cases of pneumonia/100 patients were observed. The main pathogens associated with BSI were coagulase-negative staphylococci (51%). Conclusions: The ongoing ONKO-KISS project adds to the improvement of quality of care in HCT-patients by providing sound reference data on the occurrence of BSI and pneumonia during neutropenia. Since 2006, surveillance is extended to neutropenic patients with acute leukemia to allow participation for centres not performing HCT. T-cell-mediated immunity is an essential host factor in the control of HCMV latency in patients undergoing an allohematopoietic-stem-cell transplantation. Our aims were to identify patterns of HCMV-specific immune responses associated with multiple or prolonged reactivations. We analyzed 116 recipients during the course of infection/reactivation and latency. The CD8+T-cell responses were weekly determined using HLA-class I tetramers together with extended phenotypic analyses. Our results showed that recipients from unrelated donors were more susceptible to multiple reactivations and that the donor HCMV serologic status influenced the occurrence of prolonged reactivations. We found that the lack of HCMV-specific T-cells during the first episode of reactivation was associated with multiple further reactivations. The sequential phenotypic follow up showed that patients with uncontrolled reactivations were unable to develop HCMV-specific T-cells of the late differentiation phenotype CD45RA+CD27-CD28-. Our data indicate that the longitudinal evaluation of CD27 and CD45RA expression within the tetramer positive subset could help to identify patients developing a protective immune response. The evaluation of HCMV-specific immune responses during the first episode of reactivation, together with extended phenotypes could improve immune monitoring, especially in recipients from unrelated donors and other situations at risk of uncontrolled viral reactivation.

Invasive aspergillosis in allogeneic stem cell transplant recipients from alternative donor: incidence, risk factors and outcome A.M. Raiola*, M. Mikulska , B. Bruno, M.T. van Lint, F. Gualandi, D. Occhini, A. Dominietto, C. Di Grazia, S. Bregante, A. Ibatici, T. Lamparelli, E. Furfaro, F. Frassoni, C. Viscoli, A. Bacigalupo S. Martino's Hospital (Genoa, IT) Introduction: Invasive Aspergillosis (IA) remains an important complication with high morbidity and mortality in patients undergoing haematopoietic stem cells transplant (HSCT) according the donor type. Materials and methods: We determined, with retrospective analysis, the incidence, risk factors for IA and outcome in 306 patients who received HSCT from unrelated donor (MUD), family mismatched donor, or cord blood between January 1999 and December 2006 in our Unit. The diagnosis of IA was documented as proven or probable according to the 2002 European Organisation for Research and treatment of cancer EORTC/NIAID international consensus. We have also considered proven IA the presence of fungal invasion on autopsy. Results: A total of 306 patients were included in the study, with a median follow-up of 297 days after HSCT, (range 1-2709 days) (types of HSCT: matched unrelated 60%, mismatched related 23%, mismatched unrelated 11%, cord blood 6%). There were 8 cases of proven and 37 probable IA, with prevalence of 14,7%. The median time to onset of was day + 53 (range: 4-449), with 30 (66%) cases diagnosed in neutopenia (before take o secondary neutropenia). There were 7 cases diagnosed on autopsy as a proven disease while antemortem they were classified as probable (2) or possible (4); no clinical suspicion was present in 1 patient. The diagnosis of probable IA was made by galactomannan Platelia test in 29 patients and by sputum culture in 4. In 3 patients both criteria were present. In 29 cases IA was present within lungs only, whereas 14 patients developed disseminated IA. Multivariate analysis identified the following risk factors for IA: late take of neutrophil cells (p=0.001) and steroid therapy (p=0.004). Among 45 patients with IA, 34 died and 30 deaths were related to the mould infection. Mortality was 76% (p<0.001). Multivariate analysis, among patients with IA, for overall survival identified the following risk factors: ATG use in conditioning regimen (p=0.046) steroid therapy, relapse, IgA and cholinesterase at diagnosis of IA (p=0.009, 0.03, 0.041 and <0.001, respectively). Conclusions: The prevalence if IA remains high among alternative HSCT recipients, with few risk factors for IA. Clinical and radiological presentation is highly aspecific and invasive procedures are rarely feasible. About concern survival of IA immuno deficiency at diagnosis seems to be important.

Henze (2) Minimal residual disease (MRD) quantified prior to allogeneic SCT has been shown to predict outcome in children with relapsed ALL in retrospective meta analysis. Based on these results intense discussions were started as to whether transplant procedures should be adapted according to the MRD levels pre transplant. Within the ALL-REZ BFM Group we have started a prospective trial evaluating the impact of pre-transplant MRD load in a well defined group of children who received their transplant in second or subsequent remission. Between March 1999 and July 2005, 91 children with relapsed ALL treated according to the protocols ALL-REZ BFM 96 or 2002 and receiving allogeneic SCT in 2nd (n=77) or 3rd CR (n=14) have been enrolled. MRD quantification was performed within 40 days prior to SCT by real time PCR using T-cell receptor and immunoglobulin gene rearrangements as clone-specific targets with at least 1 marker with a sensitivity of 10-4. Probability of event free survival (pEFS) and cumulative incidence of relapse (CIR) in 45 patients with MRD ≥10-4 was 0.27 (±0.07) and 0.57(±0.08) compared with 0.60 (±0.08) and 0.13 (±0.06) in 46 patients with MRD <10-4 (P [EFS, Log-Rank]=0.004; P [CIR, Gray] <0.001). Clinical and therapeutical parameters were equally distributed between both subgroups. The difference in pEFS and CIR was more prominent in intermediate risk patients S2 (n=35) compared to high risk patients S3/S4/CR3 (n=56). Thus, S2 patients with MRD ≥10-4 (n=14) showed pEFS of 0.20 (±0.12) and CIR of 0.73 (±0.15), where as patients who entered transplantation with MRD <10-4 (n = 21) had a pEFS of 0.68 (±0.12) and a CIR of 0.09 (±0.09); (P[EFS] =0.020; P[CIR] <0.001). High risk patients S3/4/CR3 who entered transplantation with a MRD load of <10-4 (n=25) showed a pEFS and CRI of 0.53 (±0.11) and 0.18 (±0.08), respectively. In contrast, pEFS and CRI were 0.30 (±0.09) and 0.50 SE (±0.09) in patients who entered transplant with higher MRD load of > MRD-4. Multivariate Cox regression analysis revealed MRD as the only independent parameter predictive for EFS (p=0.006). MRD prior to allogeneic SCT proves to be the most important risk factor for outcome post transplantation. New strategies with modified SCT procedures including conditioning regimen, graft manipulation, and GVHD prophylaxis and or post transplant intervention strategies are warranted and will be performed to improve prognosis in patients with a high risk of relapse. Acute Lymphoblastic Leukemia (ALL) remains the most common indication for unrelated donor bone marrow transplantation in children. There have been major advances in conditioning regimes; supportive care and availability of donors, all of which have made transplant a viable treatment option for patients with relapsed or high-risk disease. We retrospectively analyzed data from 371 consecutive allogeneic transplants for ALL performed at the Royal Bristol Hospital for Children from October 1987 until September 2007. 371 transplants were performed in 357 patients with ALL. Disease status at the time of transplant was complete remission(CR)-1, n=72; CR-2, n=227, CR-3, n=59 , children not in complete remission n=13.(see Table 1 ) All patients were pre-treated on the current MRC-UKALL protocols for the time period. Eighty-six patients received stem cells from identical siblings or relatives. Matched unrelated donors were used in 171 children, mismatched unrelated donors in 93, haplo-identical family donors were used in 21 children. Included in these totals are 7 umbilical cord blood donations. Cyclophosphamide and Total Body Irradiation was used as conditioning therapy with the exception of children under the age of two and children receiving second transplants. Ciclosporin A was used for graft verses host disease (GvHD) prophylaxis. Short course Methotrexate was given to mismatched and sibling donors. In the unrelated group T-cell depletion was effected using CAMPATH monoclonal antibodies. The majority of patients died due to disease relapse. The incidence of chronic graft verses host disease was low with correspondingly high Karnofsky scores. This study offers a unique opportunity to analyze data from a large number of patients consistently pretreated according to United Kingdom National Leukaemia protocols and with a uniform approach to transplant conditioning, GvHD prophylaxis and supportive care over a 20 year period. Background: Intravenous BU has been used in the two Dutch pediatric stem cell transplantation centers in various myeloablative regimens with different targets for the area under the curve (AUC) and different dosing regimens of BU (once or four times daily). We retrospectively analyzed the association between BU exposure expressed as AUC and clinical outcome. Methods: All children, transplanted between 2001-2006, receiving intravenous BU as part of a myeloablative regimen, were included. Patients were separated into quintiles based on the total AUC of four days of treatment. The association with the primary endpoints death, graft-failure or relapse, EFS and the secondary endpoints (acute Graft-versus-Host Disease grade 2-4 (aGvHD), Veno-occlusive Disease (VOD) and mucositis grade 3-4), were tested using uni-and multivariate Cox regression analysis. The lowest AUC group was used as index group. Results: 102 patients were included (46 malignant indications; 56 non-malignant indications). Median age at transplantation was 3.1 years (range 0.2 to 21 years). The overall EFS and survival were 68% and 72% respectively. In multivariate analyses a total AUC between 72.5-80 mg*h/l was associated with highest EFS (85%, p=0.029) and survival (90%, p=0.021). A lower AUC BU was significantly associated with higher incidence of relapse or graft failure, whereas in the highest AUC-percentile a high incidence of treatment related mortality (TRM) was seen. Covariates were HLA-disparity and age. HLA-mismatched donors and older children showed a significantly lower EFS, the latter mainly due to a higher TRM. aGvHD occurred in 16%, VOD in 22% and mucositis in 12% of patients. These side effects significantly correlated with a high AUC BU in combination with the addition of melphalan to the conditioning regimen. The once daily dosing versus fourtimes daily did not influence any of the results. In conclusion a total AUC of BU between 72.5-80 mg*h/l is associated with highest EFS in malignant and non-malignant disease. A lower AUC correlated with more relapse/graft failure especially in HLA-mismatched donors. A higher AUC was associated with more TRM negatively influencing EFS especially in older children. The occurrence of severe aGvHD, VOD and severe mucositis was mainly related to combination of BU and melphalan. Dosing of busulfan in children by TDM to a target of 72.5-80 mg*h/l might improve EFS.

Mechanical ventilation in a recent cohort of children after allogeneic haematopoietic stem cell transplantation: risk factors and outcome S. van Gestel*, C. Bollen, M. Bierings, J. Boelens, N. Wulffraat, H. van Vught University Medical Center Utrecht (Utrecht, NL) Introduction: In previous studies, median intensive care unit (ICU) mortality in children after hematopoietic stem cell transplantation (HSCT) was 74% (range 25% -91%). It has been suggested that ICU mortality decreased over time, but we could not confirm this in a recent meta-regression analysis (van Gestel et al, submitted) . Conclusions on mortality trends were limited, since recent outcome data from children requiring mechanical ventilation after HSCT were scarce. Objective: To assess risk factors for and outcome of mechanical ventilation in a recent cohort of children after allogeneic HSCT. Design: Retrospective chart review. Setting: A pediatric ICU and HSCT centre in a university hospital in the Netherlands. Patients and Methods: All children who received an allogeneic HSCT between January 1999 and April 2007 were included. Patients who required endotracheal mechanical ventilation for more than 24 hours were identified. Potential risk factors for the requirement of mechanical ventilation and for ICU mortality were recorded. Uni-and multivariable analyses were done to identify risk factors. Results: 175 HSCTs were performed in 150 patients. Thirtyfive patients (23% of HSCT recipients) received mechanical ventilation on 38 occasions. None of the potential risk factors was significantly associated with ICU admission. There was a trend towards an increased probability of ICU admission over the years. This suggests that transplantations were carried out in sicker patients over time. Sixteen admissions resulted in death on the ICU, giving a case fatality rate of 42%. ICU mortality was mainly associated with (persistence of) multiple organ failure on the first days of admission. None of the pre-admission transplantation characteristics significantly influenced ICU mortality. Based on their Pediatric Risk of Mortality (PRISM) scores, our patients had a higher acuity of illness on ICU admission than patients in previous studies. Six-month survival in patients discharged from the ICU was 82%. Conclusion: We found a significantly lower ICU mortality in a recent cohort of children after allogeneic HSCT compared to previous studies, even though our patients were sicker. This can most probably be explained by improvements in transplantation medicine and ICU treatment strategies. Results from our study are promising, but need to be confirmed in other, preferably multi-centre, studies.

Early and comprehensive vaccination coverage in paediatric recipients of related and unrelated stem cell transplantation following coadministration of the pneumoccoccal conjugate vaccine Prevenar™ and the hexavalent combination vaccine Infanrix hexa™ R. Meisel (1) Children undergoing allogeneic hematopoietic stem cell transplantion (aHSCT) lose protective immunity to vaccinepreventable disease and thus are at significant risk for lifethreatening infections. However, important issues on reimmunization after pediatric aHSCT remain controversial due to the lack of prospective clinical trials. In the prospective multicenter vaccination trial IKAST (NCT 00169728) a total of 77 pediatric alloHSCT recipients (median age 8.3 (1.4-17.0) years) were therefore immunized with a primary series of three monthly doses of the hexavalent tetanus, diphtheria, pertussis, poliomyelitis, Haemophilus influenzae type B and hepatitis B combination vaccine Infanrix hexa (6vCV; GlaxoSmithKline Pharma) along with the heptavalent pneumococcal conjugate vaccine Prevenar (PCV7; Wyeth Pharma). Vaccination was started at 6 months after transplantation, irrespective of immunosuppressive therapy or GvHD, with a subsequent booster dose at 18 months. Immunogenicity was analysed by assessment of antibody concentrations and adverse events were prospectively collected. Prior to immunization only 12.8% and 8.2% of patients (pts) exhibited protective antibodies towards PCV7 and 6vCV vaccine antigens, respectively. As a result of highly significant increases in mean antibody concentrations (p<0.001) protection to all PCV7 and 6vCV antigens was achieved in 85.1% and 89.8% of pts after primary immunization within the 1st year after alloHSCT, and this was independent of patient age, conditioning regimen, stem cell source, and most importantly, donor type (related vs. unrelated) and in vivo T cell depletion (all p-values>0.05). Nine months later, prior to booster vaccination, 63.8% and 83.7% of pts retained protective antibody concentrations. However, mean antibody concentrations had dropped by a factor of 1.9-4.8 (p<0.001) except for pneumococcal serotype 14, thus underlining the need for subsequent vaccination. Following booster immunization, antibody concentrations increased 2.4-19.1 fold (p<0.001) indicating robust memory responses, and 91.5% and 100% of pts achieved protective antibody levels against all PCV7 and 6vCV antigens.

Vaccination was well tolerated with no vaccine-related SAEs. Our data show that early immunization of pediatric aHSCT recipients according to our simple revaccination schedule is safe and provides early and comprehensive vaccination coverage during the first 2 years following aHSCT from both related and unrelated donors.

Treosulfan-containing regimens achieve high rates of engraftment associated with low transplant morbidity and mortality in children with non-malignant disease and significant co-morbidities B.F. Greystoke* (1), S. Bonanomi (2) (1), P. Naik (1), K. Rao (2) , N. Goulden (1) , P. Amrolia (2) , R.F. Wynn (1) , P.A. Veys (2) (1)Pendlebury Children's Hospital (Manchester, UK); (2)Great Ormond Street Hospital (London, UK) Treosulfan is an alkylating agent with both immunosuppressive and myeloablative properties, which has recently been introduced as a conditioning agent in allogeneic and autologous haemopoietic stem cell transplantation (HSCT). Early studies have been performed principally in adult patients with malignant disease in whom the drug appears to be well tolerated with a low incidence of regimen-related toxicity. We report the use of treosulfan in 32 consecutive children undergoing SCT for non-malignant disease: immunodeficiency (N=18), metabolic disease (N=9), osteopetrosis (N=4) and other (N=1). Patients received a total treosulfan dose of 36 or 42gms/m²/patient given in 3 divided doses on successive days. A range of other conditioning agents and serotherapy was administered to patients who underwent family donor SCT (N=11), or unrelated donor SCT (N=21). Bone marrow (N=17), peripheral blood (N=9) and cord blood (N=5) were used as a stem cell source and in one patient a combination of marrow and cord blood from the same sibling donor was used. One patient (3%) died of early transplant-related complications. Transplant morbidity was limited and there was no VOD other than in this patient. Mucositis was mild and many patients did not require parenteral nutrition. Nappy rash with ulceration was noted in many patients. Twenty-eight patients (87.5%) established donor cell engraftment which was full in 24 (85.7%). In 25 patients (78%) there was stable donor engraftment, including sufficient lineage specific chimerism, to correct the underlying condition. Four patients required additional transplant procedures to maintain adequate donor-derived haemopoiesis, and two patients are under consideration for further HSCT. Twentyseven patients survive with a median follow up of 417 days. There were 4 late deaths due to progression of the underlying disease, GvHD or infection. Treosulfan based conditioning regimens achieve excellent engraftment with reduced regimen-related toxicity and in children with non-malignant disease at high risk for both regimen-related toxicity and graft failure. Preliminary results indicate that in a haploidentical setting CD3/CD19-depleted grafts may be advantageous regarding engraftment and immunreconstitution. Since effector cell with potential antileukemic activity are cotransfused, such grafts may be suited in particular for patients with insufficient remission. NK cells have been shown both in vitro and in vivo to mediate positive effects regarding engraftment, GvHD, immunreconstitution and relapse and are an important part of the CD3/19 depleted grafts. We analyzed NK activity in 30 CD3/19 depleted grafts, which was low compared to fresh isolated PMNCs. Several cytokines and cytokine combinations for activation of the grafts were tested and the strongest enhancement in NK activity could be obtained with IL-15. We therefore developed a protocol for the overnight incubation of the grafts according to GMP. Stimulated grafts were transplanted seven times either at day 0 or as an immunotherapeutical approach after transplantation and were well tolerated. Median cell counts for infused cells/kg bodyweight were 46x10 6 CD56+/3-, 13x10 3 CD3+, 9x10 3 CD19+, 88x10 6 CD14+ and 3,1x10 6 CD34+ cells. No occurrence of aGvHD was seen in these patients after infusion of the cells. Enhanced NK activity after stimulation could be demonstrated in vitro against K562 or primary leukemic blasts. NK cells showed increased proliferation capacity in BrdU assay and [3H]-thymidine incorporation assays, especially after additional stimulation with IL-2 while T cells showed only a moderately enhanced proliferation against third party stimulators or with OKT3. Therefore activation of haploidentical T cell depleted grafts with IL-15 is a promising tool to enhance the activity of infused NK cells and should be further investigated in clinical trials.

Background: Development of an effective strategy for patients with EPMD. Methods: 328 patients (pts) were registered. Median age is 16.2 years (yrs) (0.4-49). Primary site was extremity in 98 pts and axial/other in 217 pts (45% in the pelvis). Metastatic spread was bone marrow (BM) only in 33 pts, bone only in 137 pts and bone & BM in 105 pts, other (mainly lymph nodes) metastatic sites in 33 pts. Tumour volume was above 200 ml in 200 pts. Six VIDE induction cycles were completed by 88%. Local treatment included surgery when possible and/or radiotherapy (Rx) as indicated. Recommended HDT was busulphan (BU) 600mg/m² and melphalan (MEL) 140 mg/m² with PSCR. Median follow up is 3 years (range, 0-7.5). Results: Partial remission or better was achieved after cycle 6 in 76%. The overall survival at 3 years is 32%±3. In patients with BM/Bone metastasis, significantly favourable univariate factors in the unselected cohort at diagnosis (Dx) were age < 14 yrs (event free survival at 3 yrs (EFS) 39%, p<0.001), metastatic site (EFS: BM only 47%, Bone only 22%, BM+Bone 15%, p=0.003), single bone lesions only (EFS 37%, p=0.004), primary tumour site (EFS: Extremities 33%, Chest/Spine/HN 27%, Abdomen Pelvic 22%, p=0.035) and tumour volume of <200ml (EFS 44%, p<0.001). ). Multivariate analysis identified four major risk factors at Dx: primary tumour volume >200ml p<0.001 (RR 2) and > 5 bone metastases p=0.055 (RR 1.7), age above 14 p=0.012 (RR 1.6), BM metastasis p=0.042 (RR=1.4). For pts receiving BuMel it is noteworthy that 46pts of <14a and EPMD achieved an EFS of 46% in comparison to older counterparts >14a (EFS 23%, p=0.007). Conclusion: Further strategy refinement and validation of HDT appears necessary within investigational , ideally randomised studies.

Allogeneic haematopoietic cell transplantation for chronic lymphocytic leukaemia with 17p Deletion: a retrospective EBMT analysis J. Schetelig*, A. van Biezen, R. Brand, D. Caballero, R. Martino, M. Itala, J. García-Marco, L. Volin, N. Schmitz, R. Schwerdtfeger, A. Ganser, F. Onida, B. Mohr, S. Stilgenbauer, M. Bornhäuser, T. de Witte, P. Dreger on behalf of the Chronic Leukemia Working Party, EBMT Purpose: Patients with advanced chronic lymphocytic leukaemia (CLL) and 17p deletion have a very poor prognosis even after intensive chemotherapy. While allogeneic hematopoietic cell transplantation (HCT) has the potential to cure patients with advanced CLL it is not known whether this holds true for patients with 17p deletion. Patients and Methods: Patients with 17p-CLL who had received HCT were identified by an EBMT-based survey. Baseline data were downloaded from the EBMT database. Additional information on the course of the disease, the cytogenetic diagnosis and last follow up was collected by a questionnaire. Data were analysed as of February 2007. Results: 56 patients were identified. Twelve patients with autologous HCT, haplo-identical donors or the detection of 17p-after HCT were excluded from further analysis. 44 patients had received an allogeneic HCT between March 1995 and July 2006 from a matched sibling donor (n=24) or an alternativ donor (n=20). The median age at HCT was 54 years. The diagnosis of deletion 17p-was made by FISH in 82% and by conventional banding in 18% of patients. The median interval between first diagnosis and detection of 17pwas 2.4 years and the median interval between detection of 17p-and HCT was 0.5 years. Patients had received a median of 3 chemotherapy regimens, including fludarabine in 98% of patients. At HCT, 53% of patients were in remission. Reduced intensity conditioning was applied in 89% of patients. 93% of the patients received peripheral blood stem cells. GVHD prophylaxis was performed heterogeneously. One patient experienced primary graft failure. Acute GVHD grades II to IV occurred in 44% of patients and extensive chronic GVHD in 46% of patients. After a median follow-up of 23 months (range, 2 to 90 months) of 24 patients who are alive, 18 were in complete remission, 4 in partial remission and 2 patients had progressive disease at last follow up. 4-year overall survival and progression-free survival was 47% (95% CI, 29% to 65%) and 38% (95% CI, 20% to 56%). The cumulative incidence of relapse at 4 years was 35% (95% CI, 12% to 57%). No additional relapse occurred in six patients with a follow-up between 4 and 7.5 years. Conclusion: Allogeneic HCT has the potential to induce longterm disease-free survival in selected patients with advanced 17p-CLL. Given the otherwise very dismal outcome of this disease, prospective studies on allogeneic HCT earlier in the course of 17p-CLL seem warranted.

Separating patients with myelofibrosis with myeloid metaplasia into risk groups for allogeneic haematopoietic cell transplantation -results of a German multicentre analysis F. Collenbusch, R. Schwerdtfeger, M. Schleuning, C. Schmid, J. Finke, M. Stadler, M. Bornhaeuser, D. Messerer-Schmid, H 

Objective: We aimed to derive risk factors for allogeneic hematopoietic cell transplantation (alloHCT) in patients with S53 myelofibrosis with myeloid metaplasia (MMM) from retrospective analysis. Patients: Between 1999 and 2006 97 patients (pts) from 15 German centers, median (md) age 51 (19-66), were grafted from 41 related and 56 unrelated donors. 74 pts had primary, 23 secondary MMM. At alloHCT 32 pts were Dupriez score 0, 37 score 1 and 28 score 2, 49 needed red cell transfusions. Chromosomal and molecular analysis was available from 75 and 51 pts resp.: 51 pts had favourable, 24 unfavourable cytogenetics, 31 pts were JAK2 mutated. Conditioning was of standard, intermediate and reduced intensity in 27, 57 and 13 cases. 28 pts received BM, 69 PBSC. 5 boosts and 16 donor lymphocyte transfusions were given for 16 relapses, 2 incomplete chimerisms and one graft failure. Results: At a md follow up of 985 (100 -2630) days probability of overall and relapse free survival (OS and RFS) was 62 and 39%. Neither time from diagnosis to transplant, age, Bsymptoms, marrow fibrosis, splenomegaly, JAK2, hemoglobin <10 g/dl, platelets <100/nl, LDH, comorbidity (CCI or HCT-CI), conditioning, type of donor nor graft source were predictive for OS. Log rank test showed a trend towards lower survival in cases with Dupriez score ≥1 (p=0.06) and transfusion dependence (p=0.1). Solely circulating blasts >1% (p=0.016), monocytes >1/nl (p=0.008) and cytogenetics (p=0.009) were predictive for survival after alloHCT. Cox regression analysis within pretransplant variables revealed cytogenetics (p=0.016) and monocytes (p=0.036) as independent factor for OS. For RFS only cytogenetics retained independence (p=0.003). Combining cytogenetics, monocytes, blasts and transfusion dependence to a risk score from 0 to 4 allowed to discriminate between good (0-1), intermediate (2) (3) and high risk (4) pts for OS (83, 45, 0%, p=0.0008) and RFS (68, 34, 0%, p=0.0013). Following alloHCT pts with limited cGvHD had a projected OS and RFS of 84% and 81% as opposed to 43% RFS for no and 20% RFS for extensive cGvHD, resp. (p=0.018). 7 pts treated with cellular immunotherapy achieved a partial and 7 a complete response. Conclusions: Pts with MMM can be scored as low, intermediate and high risk for alloHCT according to pretransplant disease characteristics. Evidence for a potent graft-versus-MMM-effect suggests exploitation of adjuvant cellular immunotherapy to improve current results.

Syngeneic BMT for CML in chronic phase: update of the Seattle results A. Fefer*, J. Radich, T. Gooley, L. Holmberg, M.E. Flowers, S. Pavletic, R. Storb, F. Appelbaum Fred Hutchinson Cancer Research Center (Seattle, US) To determine the efficacy of syngeneic BMT as treatment for chronic phase CML (CML-CP), we reviewed the results of all patients (pts) we transplanted between 1976 and 2000, before the Gleevec era. Of 34 pts, 5 died within 4 months (mos)due largely to pulmonary problems. Of the 29 pts who went into a complete cytogenetic remission (CR), 15 relapsed at a median time of 2 years after BMT, but 14 remained in CR for up to 31 (median,19) years. All 9 pts tested were negative by PCR when last tested at 2-23 (median,13) years. Thus, enduring cytogenetic and molecular CR's were achieved with syngeneic BMT, in the absence of an allogeneic graft vs leukemia (GVL) effect. At 20 years, the estimated probability of relapse, overall survival and relapse-free survival was 46%, 42% and 28%,respectively. To assess the influence, if any, of the conditioning regimen on the relapse rate, we subdivided the 34 pts into 3 groups transplanted in 3 time periods with 3 different conditioning regimens, as follows: Group I (n=14), 1976-81, Dimethylbusulfan (DMB) 5mg/kg IV, Cyclophosphamide (CY) 60mg/kg/day x2 and total body irradiation (TBI) 10Gy at a single exposure;

Group II (n=10), 1982-88, CY 60mg/kg/d x2 & TBI 2Gy/day x6; Group III (n=10), 1989-2000, Busulfan (BU) 2mg/kg/day x4 po, CY 30mg/kg/day x2 and TBI 2Gy/day x6. The median time from diagnosis to BMT for the 3 groups was 11, 3 & 5 months, respectively. Relapses occurred in 5 pts in Group I at a median of 4 years, 8 pts in Group II at a median of 1 eayr, and 2 pts in Group III at 1 and 4 years. Thus, the relapse rate with the CY/TBI regimen was significantly higher than that for the regimens containing DMB or BU (hazard rate (HR) of relapse for Group II vs I was 4.17, P=0.01, and for Group II vs III, HR 4.80, P=0.05). However, CY/TBI was less toxic, with no early BMT-related deaths in contrast to 2/14 and 3/10 deaths in Groups I and III. Finally, we compared the risk of relapse in Groups I and III combined i.e.twins conditioned with DMB/BU-containing regimens, with that of 366 CML-CP pts who underwent BMT from allogeneic matched siblings after CY/TBI at our Center. The risk of relapse is quite similar (HR=1.06, P=0.87). These results suggest that either the DMB or BU can eradicate a number of CML cells similar to that destroyed by an allogeneic GVL effect and/or that the DMB or BU can somehow induce a GVL effect exerted by syngeneic cells. We have previously shown that levels of the polycomb group (PcG) gene BMI-1 RNA were significantly higher in advanced phase than in chronic phase (CP) chronic myeloid leukaemia (CML). In addition, in patients treated with HU and IFN-a, low BMI-1 expression was associated with an improved overall survival (OS) (Blood 2007). Here, we investigated whether BMI-1 and other previously established prognostic genes (CD7, PR-3 and ELA-2) are implicated in the prognosis of CML in the context of allogeneic stem cell transplantation (allo-SCT). We studied 84 CP-CML patients who received allo-SCT from HLA-identical related donors. CD7, PR-3, ELA-2 and BMI-1 expression was assessed by Q-RT/PCR in the recipients PBMCs collected before allo-SCT. The median expression level for each gene was used to segregate the patients into 2 groups (low: gene expression <median, and high: >median). The median FU post-allo-SCT was 9.9 (range, 1.7-23.9) years. The median EBMT-Gratwohl score was 3. None of the 4 tested genes showed any significant association with engraftment or with graft rejection. CD7, PR-3 and ELA-2 expression was not associated with OS. However, in contrast to our previous findings in the non-allo-SCT setting, patients displaying a high BMI-1 expression level prior to allo-SCT had significantly better OS than those with low expression (P=0.005). When BMI-1 was included in a multivariate survival model and adjusted for the other prognostic variables, a high expression was found to be an independent marker associated with better survival (RR=2.72, 95%CI;1.1-6.9;P=0.034). Given the impact of BMI-1 expression level on OS, without a significant association with relapse, and since neither BMI-1, nor the other genes showed any significant association with leukemia-free survival, we assessed their impact on TRM. There was a striking and significant association between acute GVHD and BMI-1 expression, not only in overall incidence (low BMI-1: grade 0-1 (n=21), grade 2 (n=10), grade 3-4 (n=9); high BMI-1: grade 0-1 (n=32), grade 2 (n=9), grade 3-4 (n=1); P=0.005), but also in cumulative incidence at day 100 (48% vs. 24%, P=0.016). In multivariate analysis, a low BMI-1 expression level was associated with an increased risk of grade 2-4 acute GVHD (RR=2.85, 95%CI; 1.3-6.4; P=0.011). These results suggest that BMI-1 can serve as a biomarker for predicting outcome in CP-CML patients receiving allo-SCT. Such measurement allows for tailored therapeutic intervention, including informed recommendation for allo-SCT in patients failing tyrosinekinase inhibitors.

Haematopoietic stem cell transplantation in Tprolymphocytic leukaemia: a retrospective EBMT analysis W. Wiktor-Jedrzejczak*, R. Brand, A. Van Biesen, E. Carreras, V. Leblond, G. Cook, M. Ethell, A. Nagler, T. Ruutu, M.L. Brune, J. Schetelig, T.M. De Witte, P. Dreger on behalf of the CLWP EBMT T prolymphocytic leukemia (T-PLL) is a rare, aggressive neoplasia of T lymphoid lineage which is characterized by poor survival of less than one year. Incidental reports suggest that both autologous and allogeneic hematopoietic stem cell transplantation (HSCT) might be effective in this disease. However, no larger series on the efficacy of HSCT in T-PLL has been reported to date. Therefore the purpose of the present study was to analyze the outcome of transplants for T-PLL registered at the EBMT database. Results: Eleven patients who had undergone autologous transplantation (auto-SCT) and 33 patients who had undergone allogeneic transplantation (allo-SCT) were identified from the database and verified as T-PLL. In the auto-SCT group, there were 9 males and 2 females, with a median age of 59 (37-64) years. 5 patients were transplanted within the 1st year of diagnosis, and the remainder in later phases. 2 patients received TBI in their conditioning, while 8 received chemotherapy only. Engraftment and recovery was prompt. There was one case of non-relapse death (NRM). 5 patients relapsed within the first 2 post-transplant years, translating into a median event-free survival (EFS) of 14 months and a median overall survival (OS) of 22 months. A single patient had prolonged disease control until relapse after 53 months.In the allo-SCT group (20 males/13 females), median age was 51 (24-71) years; 18 patients (55%) had been transplanted within the 1st year of diagnosis. Status at allo-SCT was complete or partial remission in 18 patients (55%), and more advanced disease or unknown in the remainder. 14 patients (42%) received reduced intensity conditioning. Donors were HLA-identical siblings in 16 patients (48%), matched unrelated donors in 16 patients, and a mismatched unrelated donor in one patient. Engraftment was documented in 29 of 30 patients (97%) with information available. 2-year NRM was 25%. 10 of 11 relapses observed occurred during the first post-transplant year, mainly in patients transplanted in advanced disease, translating into biphasic survival curves with initial steep decline (median EFS 7 months, median OS 11 months) but still 32% EFS at 24 months suggesting sustained disease control in a subset of patients. Conclusions: These data indicate that allo-SCT may provide effective disease control in selected patients with T-PLL. Prospective investigation of allo-ST in eligible patients with this otherwise fatal disorder seems to be warranted.

High-dose chemotherapy with autologous stem-cell support versus standard-dose chemotherapy: metaanalysis of individual patient data from 15 randomised adjuvant breast cancer trials T. Demirer*, N.N. Ueno, M. Bregni, D with autologous haematopoietic stem cell transplantation (HSCT) for the treatment of primary poor risk breast cancer (BC) patients (pts) has lost favour over recent years in the oncology community. This is mainly due to the worrisome high mortality and morbidity of the procedure, and the lack of a clear survival benefit in early randomized studies. Recently reported trials have demonstrated that HDC with HSCT could still have a role in selected subgroup of pts, namely pts with ≥ 10 positive axillary lymph nodes (LN) and in pts with HER-2 negative tumours. Aim of this study is to re-evaluate toxicity and efficacy of HDC with HSCT in a large cohort of pts receiving HDC in Italy between January 1, 1990 and December 31, 2005. 1294 BC pts receiving HDC for poor risk BC were identified in the GITMO registry. In 1183 patients with >3 LN, a thorough data set including biological characteristics, toxicity and follow up was available. Median age was 46 years (24-66), 62% of pts were pre menopausal at treatment, 71% had an endocrine responsive tumours and 43% had a her-2+ tumour. Median number of positive axillary LN was 15 (4-63), with 23% of pts having ≥ 20 LN+. 73% of pts received alkylating agents-based HDC as a single procedure while 27% received Epirubicin or Mitoxantrone-containing HDC, usually within a multi-transplant program. Transplant related mortality (TRM) at 100 days was 0.7%, while late cardiac and secondary tumour related mortality were around 1% overall. With a median follow up of 74 months, median disease free survival (DFS) and overall survival (OS) in the entire population were 6.5 and 7.5 year, respectively. Exploratory subgroup analysis demonstrated that OS was significantly better in endocrine responsive tumours (p=0.0000), while menopausal or HER-2 status did not affect survival. In 85 poor prognosis pts with ER, PgR and HER-2 negative tumours (median LN+ = 18), median OS was 110 months. Median OS was significantly better (p=0.0000) in patients receiving multiple transplant procedures, this effect being particularly evident in the ≥ 10 LN+ population. In conclusion, our retrospective analysis suggests that HDC with HSCT has lower TRM than expected and high efficacy in well defined subgroups of patients. Multiple transplants seem more active than single HDC procedures. This analysis could be useful in selecting well defined patient populations in which to re-address the role of HDC as adjuvant treatment.

Long-term follow-up of metastatic renal cancer patients undergoing reduced-intensity allografting M. Bregni*, M. Bernardi, P. Servida, A. Pescarollo, R. Crocchiolo, E. Treppiedi, F. Ciceri, J. Peccatori Scientific Institute San Raffaele (Milan, IT) Objective: Stem cell transplantation from a HLA-compatible sibling donor has been advocated as adoptive immunotherapy for cytokine-resistant, metastatic renal cancer (RCC). However, the recent introduction of several targeted therapy compounds has reduced the interest in this therapeutic strategy. We have reanalyzed our transplant series with the aim to detect long-term benefit form allografting. Methods: from February 1999 to May 2005, 25 patients with cytokine-refractory RCC received a reduced-intensity allograft from an HLA-id sibling donor. Median age was 53 years; most (24) had clear-cell histology. Median number of previous treatments was 1 (0-3). Median days from diagnosis to allograft were 822. All patients received a thiotepa, fludarabine, and cyclophosphamide conditioning regimen, and a cyclosporine-based GVHD prophylaxis. Six patients received DLI at escalating doses for progressing or nonresponding disease. Results: One-year-OS was 48% (95% CI: 28-68), and 3y-OS was 20% (95% CI: 4-36). At a median observation time of 65 months, 5 patients are alive, one in CR, one in VGPR, and three with disease. We have analyzed the correlation of the following variables with survival: pre-transplant disease status, age at transplant, time from diagnosis to transplant, total infused cells, Infused CD34+ cells, Infused CD3+ cells, BM chimerism at +30, post-transplant disease status at +30 and +90, best response, day of best response, CSA withdrawal day, infusion of DLI, occurrence of GVHD. At univariate analysis, numbers of CD34+/kg infused cells, best response, and disease status at +90 significantly correlated with survival. At multivariate analysis, only disease status at +90 retained statistical significance (p=0.002), while CD34+ cells/kg retained marginal significance (p=0.059). Conclusion: Twenty percent of cytokine-refractory RCC patients are alive at a median 65 (40-72) months after allografting. All these patients have received >5x10 6 CD34+ cells/kg, and had stable or responding disease at +90 after transplant. Three patients are receiving sorafenib, and two are in CR/PR without further therapies. Reduced-intensity allografting is able to induce long-term disease remission in a fraction of relapsed RCC patients. It is unknown if relapse or PD after targeted therapy will be susceptible to allograft-mediated GVT effect. The place of allografting in the treatment of metastatic RCC, alone or in combination with targeted therapies, needs reappraisal. Haploidentical family donors represent the ideal solution to offer to every patient with high risk leukemia the potential cure of marrow stem cell transplantation. Extensive application of haploidentical transplantation (haplo-SCT) is limited by high rate of late transplant related mortality (TRM) and relapse associated with the delayed immune reconstitution secondary to the procedures for severe graft-vs-host-disease (GvHD) prevention. In a haplo-SCT phase I-II multicenter, open, no-randomized, trial sponsored by MolMed SpA, we infused donor lymphocytes genetically engineered to express the suicide gene herpes simplex thymidine kinase (TK-DLI) to induce early immune reconstitution, while selectively controlling GvHD. Between September 2002 and September 2007, 51 patients (pts) -median age 48-with high-risk hematologic malignancies were enrolled, 29 out of 51 pts were in complete remission (CR). After myeloablative conditioning regimen, 48 pts received a median 13x10 6 /kg CD34+ and 1.0x10 4 /kg CD3+ (median time to engraftment: 2 weeks). No immune reconstitution were observed in absence of TK-DLI. Twenty-seven pts received TK-DLI: 22 pts obtained prompt immune reconstitution with CD3+>100/mcl at day+75 (median) from haplo-SCT and day+23 from TK-DLI. Eleven pts developed GvHD (10 acute GvHD grade I-IV and 1 chronic GvHD) that was always abrogated by the suicide gene induction. The 3-year TRM was 26%, with last infectious event at day 166 post transplant, for pts treated with TK-cells, and 69% for pts who didn't receive TK-cells (p<0.0001). Immune reconstitution obtained with TK-cells infusion correlated with: 1. rapid development of a wide T-cell repertoire, 2. detection of high frequencies of T-cells specific for opportunistic pathogens, 3. abatement of the incidence of infectious adverse events (AE) and serious AE. The 3 years LFS was 45% for pts who achieved immune reconstitution and 9% for pts who failed immune reconstitution (p:<0.0001). This strategy is feasible and effective in providing immune reconstitution in haplo T-cell-depleted setting. In uni-and multi-variate analysis both status at transplant and immune reconstitution are significant risk factor. Infusion of TK-cells could significantly extend the application of haplo-SCT. A randomized phase III study comparing TK-DLI versus any T cell repletion strategy after haplo-HSCT in high risk acute leukemia is now starting.

IE-1 and pp65 specific peptide pools for the generation and expansion of CMV-specific donor T-cells after haploidentical stem cell transplantation: feasibility and first clinical experiences S. Ganepola* (1), P. Reinke (2) , C. Gentilini (1) , M. Hammer (2) , M. Schmidt-Hieber (1), C. Tietze-Bürger (1), K. Freyberg (1), D. Volk (2) , E. Thiel (1) , L. Uharek (1) (1)Hematology/Oncology/Transfusionmedicine (Berlin, DE); (2)Charite Berlin Campus Mitte (Berlin, DE) For immunocompromised patients, rapid reconstitution of cytotoxic T cell function is mandatory for the control of human cytomegalovirus (CMV) infection and disease. The CMVassociated proteins pp65 and IE-1 are important components of a relevant CMV-directed immune response. Here we report the first clinical experiences concerning feasibility and safety with a novel method for the generation of CMV-specific T-cells by stimulation with a peptide-pool containing pp65 and IE-1. Material and Methods: After apherisis of 1-1.2x10 9 PBMC from 4 healthy donors (2 IgG positive, 2 IgG negative) and 2 patients with proven CMV disease, cells were further processed under GMP-conditions. Cells were stimulated with peptide pools representing the pp65 (UL83) and IE-1 (UL123) proteins and IFNg producing cells were selected with the cytokine-capture-assay (MiltenyiBiotec) and further expanded in cell-culture conditions on a 24-well-plate. After performance of intern and extern quality controls, cells were freshly retransfused and/or kryoconserved in defined portions for further redonations. Results: In 5/6 rounds we could expand CMV-specific T-cells. Expansion-rates ranged from 1.2 fold to 71.1 fold (median 9.4) of the initial cell count. In a median time of 15 days (range 14-23 days), cell counts expanded from 1.2 fold up to 71.1 fold (median 9.4). FACS-analysis at the end of the culture revealed that in median 80% of the cells were CD8+ (range 2.5-92.3%) and, respectively, 20% were CD4+ (range 7.2-89.6%). In cytotoxicity assays, a lysis of 50-83% (median 53%) of LCL-line cells could be achieved. So far, two patients with therapy refractory CMV disease received in vivo expanded polyvalent T-cells against multiple (IE-1/pp65) CMV epitopes. No serious adverse events occurred during the first days after administration and CMV antigenemia was decreasing in one patient. However, since both patients finally died from a septical multi-organ failure, we were not able to assess the long lasting impact of the manoeuvre on the control of CMV-disease. Conclusion: Selection and subsequent expansion of CMVspecific T-cells is possible with this method which allows an up to 70-fold expansion of CD8+ cells. The use of pp65 and IE-1 specific peptide pools for the prevention or treatment of CMV infection is feasible and could be of superior effectiveness as compared to approaches directed against a single CMV epitope.

LMP2-specific TCR gene therapy for Hodgkin's lymphoma D.P. Hart, S. Thomas, S. Xue, H.J. Stauss, E.C. Morris* University College London (London, UK) EBV-positive Hodgkin lymphoma typically demonstrates latency II antigen expression, characterised by loss of most EBV antigens except for the latent membrane protein (LMP) 1 and 2 and the EBNA-1 protein. Reed Sternberg cells expressing LMP2 can be a target for antigen-specific immunotherapy, but LMP2-specific autologous CTL for adoptive immunotherapy are difficult to generate due to poor immunogenicity. T cell receptor (TCR) gene transfer using retroviral vectors containing the TCR alpha and beta chain genes can reproducibly redirect the antigen specificity of a T cell population. The aim of this study was to generate a retroviral TCR construct suitable for the rapid and efficient production of LMP2-specific CTL. Retrovirally introduced TCRs compete with endogenous TCRs for CD3 molecules required for assembly of the TCR complex. This competition may limit surface expression of the introduced TCR resulting in a transduced T cell with poor functional avidity. In an attempt to generate a 'highly competitive' LMP2-TCR the following modifications were made to the retroviral vector construct: i) nucleotide sequences were codon-optimised for efficient translation in human cells; ii) the constant region of each TCR chain was altered to contain murine sequences to enhance CD3 binding; and iii) the TCR alpha and beta chain genes were linked by a self-cleaving 2A sequence. The unmodified HLA-A2-restricted LMP2-specific TCR was poorly expressed in primary human T cells (up to 2.5% of viable CD3+ T cells, as detected by FACs analysis using monoclonal anti-Vbeta13 antibodies), suggesting that it competed inefficiently with endogenous TCR chains for cell surface expression. However, retroviral transfer of the modified LMP2-TCR into human T cells improved LMP2 TCR expression to 55-65% CD3+ T cells. The transduced cells bound HLA-A2/LMP2 pentamer, showed peptide-specific IFNgamma and IL2 production and killed target cells displaying the LMP2 peptide. Importantly, expression of the introduced LMP2-TCR suppressed expression of almost the entire repertoire of endogenous TCR combinations, including 'mis-paired' TCRs. 'Mis-paired' TCRs contain an introduced alpha chain paired with an endogenous beta chain and vice versa. The antigen specificity of such mispaired TCRs is unknown and could be auto or allo-reactive. Modified TCR sequences, producing 'dominant' TCRs may improve the efficacy and reduce the potential risks of TCR gene therapy a novel form of adoptive immunotherapy.

Allogeneic stem cell transplantation in patient with Major Histocompatibility Complex class II immunodeficiency: a single-centre experience H. Al-Mousa*, Z. Al-Shammari, A. Al-Ghonaium, H. Al-Dhekri, S. Al-Muhsen, R. Arnaout, A. Al-Seraihy, A. Al-Jefri, A. Al-Ahmari, M. Ayas, H. El-Solh King Faisal Specialist Hospital (Riyadh, SA) Background: Major Histocompatibility Complex Class II (MHC II) deficiency is a rare combined immunodeficiency disease. Allogeneic bone marrow transplantation (BMT) is considered the only available curative treatment. Survival rate post BMT is lower than other forms of primary immunodeficiencies. These differences were observed for both BMT performed with matched and non-matched donors. Patients and methods: Between June 1994 and august 2007, thirty two children with MHC II deficiency underwent thirty seven BMT procedures at King Faisal Specialist Hospital and Research Centre, Riyadh, Saudia Arabia. Five patients required second BMT trial. Median age at BMT was 27 months (range, 1-129 months). The source of stem cells was unmanipulated marrows from HLA-genoidentical siblings in 22 pts, HLA-phenotypically identical related donors in 9 pts and umbilical cord for 1 patient. Conditioning was with one of 4 regimens, regimen A (19 pts): Busulfan (BU), Cyclophosphamide (CY) and Etopside (VP-16), regimen B (3 pts): BU/CY and anti-thymocyte globulins (ATG), regimen C (2 pts): BU/CY/VP-16 and ATG and regimen D (13 pts): fludarabine, melphalan and ATG. Median CD34 dose was 8.3 x 10 6 /kg (range, 1.5-20.7 x 10 6 /kg). Graft-versus-host disease (GVHD) prophylaxis consisted of cyclosporine (CSA) and methorexate (MTX) in 21 pts, CSA alone in 15 pts and CSA and steroid in 1 patient. Results: 24 pts had adequate immune reconstitution and sustained engraftment (assessed by short tandom repeats) ranged from 24-100% for lymphoid line and 5-100% for myeloid line. Seven pts (4 pts were post second transplant) died secondary to sepsis, multiorgan failure and primary disease. Acute GVHD was seen in 20 pts and 4 pts developed chronic GVHD. The overall disease free survival rate was 75% with a median follow up of 5.3 years (range 0.5-10.2 years). Low survival rate (20%) was seen in patients who underwent second BMT. Conclusion: Bone marrow transplantation (BMT) can cure the disease, provided it is performed before complications leading to severe organ failure develop. Second BMT is associated with high rate of mortality. Further studies and long term follow up are required to determine the appropriate conditioning regimen. To analyze the effects of CD34-TRAIL+ cells on tumor vasculature, tumorbearing mice were perfused with sulfo-biotin and tumor endotelial cells (TEC) were then revealed by horseradish peroxidase (HRP)-conjugated streptavidin. As compared with CD34-mock-or soluble (s)TRAIL-treated mice, a 24-hour treatment with CD34-TRAIL+ cells significantly (P ≤.001) reduced microvessel density (1850 ± 1139 vs 2227 ± 915 vs 757 ± 562 vessel per 1 x 10 5 tumor cells, respectively) and increased the thickness of the vessell wall (3.7 ± 1 µm vs 3.4 ± 1 µm vs 6 ± 1 µm, respectively), suggesting that CD34-TRAIL+ cells induce an early vascular disruption leading to a progressive disintegration of the vascular bed. Confocal microscopic imaging of tumor sections double-stained with anti-CD31 and anti-TRAIL-R2 showed that this receptor was expressed by 8 -12% of large tumor vessels. Interestingly, upon treatment with CD34-TRAIL+ cells, but not sTRAIL, TUNEL staining revealed an extensive apoptosis of TEC. Forty-eight hours following injection of CD34-TRAIL+ cells, a 21-fold increase of apoptotic index was detected, which was associated with extensive necrotic areas (20% to 25% of tissue section). These data show that: (i) tumor homing of CD34-TRAIL+ cells induces extensive vascular damage, hemorrhagic necrosis and tumor destruction; (ii) the antitumor effect of CD34-TRAIL+ cells is mediated by both indirect vascular-disrupting mechanisms and direct tumor cell killing.

Targeting of a therapeutic suicide gene to human alloreactive memory T-cells with stem-cell features requires IL-7 and IL-15 A. Bondanza* (1), L. Hambach (2) In a phase II clinical trial investigating the prophylactic infusion of suicide gene-modified donor T cells after haploidentical hemopoietic cell transplantation (haplo-HCT), we observed a rapid and effective immune reconstitution. After activation with anti-CD3 antibodies, T cells were modified with a retroviral vector (RV) encoding for the Herpes Simplex thymidine kinase (TK). TK+ cells displayed an effector memory (EM) phenotype (CD45RA-CD62L-, CD28±CD27+, IL-2±IFN-g+). When needed, graft-versus-host disease (GvHD) was controlled upon administration of the prodrug ganciclovir (GCV). The graft-versus-leukemia (GvL) effect was substantial in patients transplanted in remission, but failed to cure patients in relapse. Genetic modification with RV is limited to memory T cells. EM TK+ cells have a reduced alloreactivity. Central memory (CM) T cells (CD45RA-CD62L+, CD28+CD27+, IL-2+IFN-g±) share many characteristics with stem cells, namely the ability to selfrenew and to differentiate into effector cells. Recently, it has been proposed that alloreactivity may be confined to memory T cells with stem-cell features. Since alloreactivity is the common ground of both graft-versus-host disease (GvHD) and the GvL effect, crucial to the success of the strategy is the suicide gene-modification of cells with such properties. We found that addition of CD28 costimulation on cell-sized beads and the use of homeostatic cytokines, such as IL-7 and IL-15, generates central memory (CM) TK+ cells. CM TK+ cells were highly alloreactive, both in vitro and in vivo in a humanized animal model of GvHD based on the grafting of human skin onto NOD/scid mice. GCV administration abrogated GvHD. Stimulation of CM, but not of EM TK+ cells with autologous dendritic cells pulsed with HLA2-restricted peptides from the minor histocompatibility alloantigen (mHag) HA-1 or H-Y efficiently induced mHag-specific T cells that lysed natural ligand expressing HLA-A2+ targets. A fraction of mHag-specific TK+ cells expressed IL-7Ra. Only IL-7Ra+ mHag-specific TK+ cells could self-renew and differentiate into effector cells. When infused in NOD/scid mice harboring human mHag+HLA-A2+ leukemia, TK+ mHag-specific T cells significantly delayed disease progression. Altogether, these data suggest that targeting of a suicide gene to human alloreactive memory T cells with stem-cell features requires IL-7 and IL-15 and warrant their use in the clinic for a safe and powerful GvL effect.

Loss of FOXP3 expression after in vitro expansion of human CD4+CD25+CD127-regulatory T-cells P. Hoffmann* (1), T.J. Boeld (1) , R. Eder (1) , J. Huehn (2) , S. Floess (2) 

In animal models the adoptive transfer of donor-type CD4+CD25+ regulatory T cells (Treg) protects from graftversus-host disease (GVHD) after allogeneic stem cell transplantation (SCT), as shown by us and several other groups. Exploring this strategy in human SCT, we currently perform a first phase I clinical trial using freshly isolated Treg. For future trials requiring large cell numbers for repetitive treatments, we described in vitro culture conditions that permit a more than 3-log polyclonal expansion of Treg (Blood 104:895; 2004) . A highly enriched starting population proved to be crucial for the generation of pure Treg cell products, a criterion fulfilled by the naïve, CD45RA+ subpopulation of CD4+CD25high T cells (RA+ Treg) (Blood 108:4260; 2006 ). An alternative isolation strategy for Treg relies on the exclusion of CD127+ cells, as activated CD4+CD25+ conventional T cells express high levels of CD127 while CD4+CD25+ Treg show no or only weak expression. For a direct comparison of the two isolation protocols, we sorted CD4+CD25+CD127low/neg T cells (CD127-Treg) and RA+ Treg from the same leukapheresis products and analyzed the cells after 2 and 3 weeks of expansion. Whereas both populations were > 94 % FOXP3+ upon isolation, only RA+ Treg maintained FOXP3 expression throughout the expansion period (93 % [range: 78 to 97 %; n=11] after 2 and 87 % [range: 71 to 97 %; n=9] after 3 weeks). In contrast, CD127-Treg cultures contained only 82 % (range: 56 to 96 %; n=11) FOXP3+ cells after 2 weeks and highly variable and significantly lower numbers of FOXP3+ cells than RA+ Treg cultures after 3 weeks (57 % [range: 18 to 93 %; n=9; p=0.006]). Further analysis identified CD45RA-FOXP3+ memory-type cells within the CD127-Treg starting population as a major source of FOXP3-as well as IL-2-and IFNgamma-producing cells emerging during in vitro culture. In addition, we analyzed the DNA methylation status of a defined region within the FOXP3 locus termed TSDR (T reg-specific demethylation region), which has been shown to be demethylated exclusively in natural T reg cells (Baron et al., EJI 37:2378; 2007) . Upon isolation, RA+ and CD127-Treg showed complete demethylation of the TSDR, but only RA+ Treg maintained this demethylated status throughout the entire culture period. Based on these findings, we suggest that expansion of naive CD45RA+CD4+CD25high T cells represents the most promising strategy for adoptive Treg cell therapies. Several EORTC AML trials were based on a 2x2 factorial design. In elderly patients, AML-13 trial was set-up to assess the value of G-CSF during/after induction, and of infusional vs non-infusional mini-ICE as consolidation. In the EORTC-GIMEMA AML10 study (patients ≤ 60 years old), comparing 3 anthracyclines in induction and consolidation, the value of alloSCT vs autoSCT has been assessed based on intent-totreat analysis. The "donor vs no donor" comparison was performed, overall, and according to cytogenetic subgroups and age. In the EORTC-GIMEMA AML12 (age ≤ 60 yrs) study the accrual for the 1st question (HD-Ara-C vs SD-Ara-C in induction) had to be expanded, due to insufficient number of patients randomized for the 2nd one (IL-2 vs No IL-2 after autoSCT). New statistical considerations were set-up for the 1st question as the trial expanded to 2000 patients.

302 Statistical planning of clinical trials: an emphasis on sample size determination A. Latouche* EBMT (Paris, FR) An essential step when planning a trial is the calculation of the sample size or the number of patients to recruit to detect a relevant effect with sufficient power. Patients enrolled in a clinical trial may experience exclusive failure causes, which defines a competing risk setting. The main decision regards the relevant quantity to assess the treatment effect: either cause-specific hazard or cumulative incidence function. The analysis should then be performed according to the method used for sample size computation, if one does not want an under-powered trial. From practical point of view, the practitioner has different options to plan a study accounting for competing risks. In this work, we thus compare these approaches.

In medical applications multi-state models arise in attempting to understand complex disease or treatment processes. In these models individuals can move among a finite number of states defined by specific conditions of health, often including death. They are natural extensions of the traditional survival models in which transitions between multiple states are considered. Multi-state models give more insight into the disease-recovery process because they take into account the occurrence intermediate events. In this talk I discuss some of the advantages (and disadvantages) of multi-state models and their application to bone-marrow transplantation.

Long-term follow-up of HCV infected patients; updated results of the EBMT prospective study P. Ljungman*, A. Locasciulli, A. Békássy, L. Brinch, I. Espigado, A. Ferrant, I. Franklin, V. Gomez-Garcia de Soria, J. O'Riordan, M. Rovira, P. Shaw, H 

Many long-term survivors after SCT are infected with HCV. It has been reported that a high proportion of these patients develop late occurring liver cirrhosis. Treatment options include interferon with or without the addition of ribavirin but the knowledge about these options' effectiveness is limited. Therefore in 1992, the IDWP of the EBMT initiated a prospective study of HCV infected SCT patients having survived at least 6 months after SCT. The intention is to every five years ask for follow-up on living patients. 247 patients were included in the study cohort between 1992 and 1997. This is the 3rd report of this cohort. Due to limited follow-up data, 49 patients have been excluded. Thus, the report is based on 198 patients with a median follow-up of 16 years (0.52 -32.8). The Kaplan-Meier estimate of survival is 73.9%. 43 patients have died of whom 7 have been reported to have died from liver associated complications including one after liver transplantation. Two additional patients are alive after liver transplantation. 79 patients have been treated with interferon with or without the addition of ribavirin or with ribavirin alone. 37/79 patients became HCV PCR negative of whom 3 have relapsed with positive HCV PCR. The side effect profile was similar to other patient populations treated for HCV. Conclusion: In this prospectively followed cohort, the prognosis of chronic HCV infection was quite favourable. Treatment is feasible and associated with acceptable efficacy and toxicity.

Current status of the APBMT registry Y. Kodera*, A. Yoshimi, R. Suzuki, Y. Atsuta, L.L. Chan, A. Li, P.-L. Tan, W.Y.K. Hwang, T.V. Binh, T.J. Chiou, A. Ghavamzadeh, L. Dao-Pei, P.- The data was submitted through the national registries in MY, JP, and TW. In other countries/regions, the data was collected either by APBMT data center or by regional coordinators. The total number of HSCT has been steadily increasing in most of the countries/regions and most dramatically in CN and in IR. The annual number of HSCT of 2005 reported was 4,598 (data of 7 countries/region), which was doubled in these 10 years; of those, 65% were allogeneic and 35% were autologous. The number of allogeneic SCT has been steadily increasing year after year; meanwhile the number of autologous SCT has been recently stable. The distribution of related and unrelated donors (UD) differed widely among the countries (UD= 0% [VN] to 62% [JP] in 2005) . UD-SCT is increasing consistently in CN, JP and HK, but not in other countries/regions. Interestingly, cord blood appeared to be a common stem cell source in Asia, accounting for 35% of UD-HSCT (11% in CN to 100% in IR/VN, data of 2005). Unrelated donor peripheral blood stem cell transplantation is not common in Asian countries/regions except for China. In summary, this simple survey provided us basic information on current situations and trends of HSCT in Asia, which may be helpful to advance the patient registry. Recently APBMT developed an electric data collecting system called TRUMP and the group also agreed to share the common basic survey items with EBMT and CIBMTR (MED-A/TED) and join the global transplant activity survey under the umbrella of World-Wide Group for Blood and Marrow Transplantation (WBMT). These international collaborations may facilitate future advances of the HSCT. Background: EBV associated PTLD is a serious complication after SCT.Several risk factors may increase the risk of PTLD. Analysis of EBV DNA viral load has been suggested to be useful for monitoring and used as the basis for preemptive therapy with rituximab with the aim to reduce the risk for PTLD. Methods: Patients who underwent SCT from 030101 until 070515 were included in this analysis. EBV serological mismatch between donor and recipient, primary EBV infection, cord blood grafts, and diagnosis of lymphoma were regarded as risk factors and used to split the patients into a high risk and a standard risk group. Before 050701 (early group), patients were tested on clinical suspicion of EBV associated symptomatic infection while after 050701 (recent group) high risk patients were monitored by quantitative PCR while standard risk patients were to be sampled on clinical suspicion of symptomatic EBV infection only. 274 patients were analyzed; 150 patients in the early and 124 in the recent group. In the early group 53 were classified as high and 97 as standard risk. In the recent group, 41 were defined as high and 83 as standard risk. The EBV viral load was measured in serum by a TaqMan PCR technique. After 050701, rituximab was to be given in high risk patients at an EBV DNA level of >10.000 copies/ml. In the early group and in standard risk patients of the recent group, rituximab was given on suspicion of EBV symptomatic infection. Results: A total of 971 blood samples were analyzed for EBV DNA. More samples (median 4 in the early and 6 in the recent group) were taken in the high risk compared to the standard risk group (median 2 and 3 samples, respectively). In the early group, 25/53 (47%) high risk patients compared to 31/97(32%) standard risk patients had EBV detected at least once. The corresponding numbers in the recent group were 14/41(34%) in the high risk compared to 11/83 (13%) in the standard risk group. In the early group, 13/150 (8.6%) received rituximab. while 11/124 (8.9%) in the recent cohort was given rituximab. High risk patients received more often rituximab than standard risk patients in both the early and recent cohorts (data not shown). In the early group 6 (4%) developed PTLD of whom 4 died (2.6%) while in the recent group, 4/124 (3.2%) developed PTLD of whom 2 died (1.6%). Conclusions: A strategy of targeted monitoring to high risk patients can be safely utilized with a low risk for development and fatal outcome of PTLD.

A retrospective EBMT survey on the use of cidofovir for BK-related haemorrhagic cystitis after allogeneic haematopoietic stem cell transplant S. Cesaro* (1), Y. Koc (2) BK virus has been recently associated to post-engraftment HC. Cidofovir (CDV) is often used for BK-related hemorrhagic cystitis (BK-HC) treatment although data on safety and efficacy are scarce for this indication. We collected retrospectively the experience of CDV for BK-HC among the EBMT centres. 61 episodes of BK-HC in 60 patients from 16 centres were recorded. All patients but 5 had been transplanted for a malignant disease. Median age at haematopoietic stem cell transplant (HSCT) was 21 yrs (range 2-64) and 23/60 patients were children or adolescent (age < 18 yrs). The source of stem cell was PB in 52%, BM in 29%, and CB in 19% whilst the type of donor was sibling in 30%, unrelated donor in 57% and other related donor in 13%. Myeloablative regimen + TBI was used in 74% of patients. BK-HC occurred after a median of 41 days from HSCT (range 3-577) but only 8 episodes were diagnosed after day + 100. HC was scored as follows: 7 grade I, 16 grade II, 28 grade III, 10 grade IV, and lasted in median 33 days (range 5-749). Concurrent morbidity was represented by fever in 30%, hypertension in 26%, VOD in 7% and TTP in 3% of episodes. Most of patients received hyperhydration plus RC and PLT transfusion during HC whilst only 51 of episodes were treated with bladder irrigation through vesical catheter. CDV was started a median of 6.5 days after HC (range -22 before to 135 days after HC) and was administered for a median of 3 doses (range 1-15). The most frequent schedule was 3-5 mg/kg/weekly or fortnightly with probenecid (used in 70% of all treatments). 6 patients received intravesical CDV via catheter. CDV toxicity was reported in 20% of episodes and was limited to kidney: 3 grade I, 4 grade II, 2 grade III, 3 grade IV. Complete resolution or clinical improvement of HC was reported in 35 (57%) and 8 (13%) of episodes, respectively. Although the datum is not available for all episodes CDV treatment was associated to negativization of BK viremia in 69% (27/39) and of BK viruria in 37.5% (9/24) of episodes. We conclude the CDV therapy is a potential useful option for the treatment of BK-HC with limited moderate or severe kidney toxicity. Prospective data are needed to define the best timing and schedule of treatment and the predictive factors for clinical and virological response.

Single-centre prognosis analysis and validation of the Seattle, French and Strasbourg prognosis indexes of invasive aspergillosis in adult patients with haematological malignancies or after haematopoietic stem cell transplantation R. Parody*, R. Martino, F. Sanchez, J.L. Piñana, J. Sierra Sant Pau Hospital (Barcelona, ES) In this retrospective study we analyzed the outcomes of 117 adults hematological patients who had a history of proven (n:23), probable (n:61) and possible (n:33) invasive aspergillosis (IA) between January 1995 and June 2007, at Santa Creu i Sant Pau's Hospital of Barcelona. Forty-one patients (35%) were recipients of an allogeneic hematopoietic stem cell transplantation (alloHSCT). The main goal of the study was to analyze the prognosis factors predicting the outcome of 4-month aspergillosis free survival (AFS) and overall survival (OS), in order to compare the results with previously published prognosis indexes. Analysis was performed in all patients and separately in AlloHSCT and non-AlloHSCT patient cohort. A total of sixty seven patients (57%, 30 recipients of AlloHSCT)) died in the first 4 months after IA diagnosis [median days: 20 (range1-117)]. Invasive aspergillosis was identified as the main cause of death in 51 patients (43%, 23 recipients of AlloHSCT) [median days: 16 (range ]. According with autopsy findings, 27 cases (11 possible and 16 probable) could be recategorized as proven cases. Diagnosis of IA at or before 2000 had a negative impact in both 4-month AFS and 4-month OS. In all patients and allo-HSCT patients, 4 variables (excluded the year of diagnosis) decreased 4-month AFS in multivariate analysis: (i) disseminated IA and monocytopenia <0.1 x 109/l, respectively (ii) impairment of one organ, (iii) impairment of 2 or more organs and (iv) high-doses steroids and an alternative donor, respectively. A risk model for progression was generated for each group, according with the presence of 0-1 factors,( 82 and 100% AFS), 2-3 factors, (27 and 45% AFS) or 4 factors, (0% AFS)[P<0.001]. In base of similar results previously published our Prognosis Index model could be validated with both the French and Seattle models for allo-HSCT recipients and with the Strasbourg model for hematological patients.

One-year microbiological survey with molecular typing method of Pseudomonas aeruginosa in a BMT Unit B. Bartolozzi* (1), R. Fanci (1) Pseudomonas aeruginosa is one of the most common nosocomial pathogens in intensive care and in oncohematological units and it still represents an important cause of morbidity and mortality. Molecular epidemiological survey is a very important tool to understand the nosocomial pattern of each hospital and to identify outbreaks. From May 2006 to June 2007 we performed a "real time" AFLP (amplified fragment-length polymorphism) monitoring of all P.aeruginosa strains isolated in patients within the BMT Unit of Florence. The patients admitted in the Unit during this period were 99 (70 autologous transplants, 15 unrelated allogeneic transplants and 14 identical sibling allogeneic transplants. P.aeruginosa was responsible for ten bloodstream infections (BSIs) and of two microbiologically documented infections without bacteremia. The AFLP analysis showed the persistence within the Unit of two clones of P.aeruginosa.

One was responsible of two outbreak episodes in patients allocated in the same single room. The first episode occurred in May-October 2006 involving 4 patients; a team consisting in the BMT head physician, the BMT head nurse, the infection control officer and the infection control physician was actively involved in a strict bacteriological surveillance that allowed to isolate the same strain from a Irgasan soap sample. For this reason, hyperclorination of the water network, sink tap changing, water filters installation and weekly soap monitoring were performed. Nevertheless the same clone was isolated after several months in a water sample, showing the inefficacy of the control measures taken. In May-June 2007 the same P.aeruginosa clone was isolated in three infected patients and in a shower tap sample. Hydraulic works inside the room were performed and until now no other case of P.aeruginosa was observed. The other persistent clone was responsible of contamination of 4 sample of Irgasan soap taken from the ward cloakroom (collected from November 2006 until April 2007) and of a BSI in a patient (March 2007) . Our results showed that P.aeruginosa is largely and consistently found in environment which may represent an important source of infection, difficult to eradicate. Moreover, punctual microbiological surveillance and molecular typing methods are essential to early detect nosocomial outbreaks, to identify P.aeruginosa reservoir and to guide in decision making in order to understand and possibly stop the epidemic chain. Donor lymphocyte infusion (DLI) following T cell depleted allogeneic stem cell transplantation (SCT) is an effective treatment for relapsed hematological malignancies. DLI often results in a profound graft versus leukemia (GVL) effect with relatively limited graft-versus-host disease (GVHD). Both GVL reactivity and GVHD are likely to be mediated by donor derived T cells recognizing polymorphic minor histocompatibity antigens (mHag) on patient cells. It has been hypothesized that T cells recognizing mHags expressed on non-hematopoietic cells are responsible for GVHD, whereas T cells recognizing hematopoiesis restricted mHags may be selectively involved in GVL reactivity. To analyze the diversity of alloreactive T cells involved in GVL and GVHD we analyzed the immune response in a patient responding to DLI following HLA-matched SCT for AML. Six weeks after DLI, GVHD limited to skin and mouth developed coinciding with a rapid and sustained conversion to 100% donor chimerism. We clonally isolated and characterized activated HLA-DR expressing T cells at the onset of GVHD (10% of the circulating T cells). In total 133 CD8+ T cells clones and 241 CD4+ T cell clones were tested for alloreactivity as defined by cytotoxicity or IFNg production upon recognition of patient derived EBV-LCL and not donor derived EBV-LCL. 49% of the CD8+ T cell clones and 20% of the CD4+ T cell clones were identified as alloreactive T cell clones. Next, recognition of patient skin-fibroblasts as a target for GVHD was determined both in the absence or presence of IFNg to mimic an inflammatory environment. None of the T cell clones recognized unmanipulated fibroblasts. Only after co-culture with IFNg, 28% of the alloreactive CD8+ T cell clones but none of the CD4+ T cell clones reacted with patient fibroblast. By blocking studies, panel studies and T cell receptor-VB analysis the diversity of this immune response was determined. 32 and 25 different reactivities were found for the CD8+ T cells and CD4+ T cells, respectively. In conclusion, isolation of activated T cells identified a very polyclonal immune response directed against multiple mHags in all HLAclass I and HLA-class II alleles. Despite the polyclonality of this immune response, reactivity of the T cell clones was relatively hematopoiesis specific, resulting in 100% donor chimerism, persistent complete remission and only moderate GVHD limited to the skin. Since T cell recognition of fibroblasts was observed only upon co-culture with IFNg, this may reflect a secondary reactivity induced by the ongoing GVL effect. Background and aims: Chronic Graft-versus-Host Disease (cGvHD) is a recognized cause of genital complications in the vulva and vagina in female patients (pts) after allogeneic stem cell transplantation. However, clinical features and consequences of genital cGvHD are poorly described in the literature. In a retrospective study, we assess prevalence, symptomatology and effect on sexual life of genital cGvHD. In a prospective study, we try to prevent progression and sequelae of genital cGvHD. Here we report early results from these ongoing studies. Patients and methods: All women allografted 1996-2005 in the Western region of Sweden are asked to participate in a study comprising (i) structured anamnesis and validated questionnaires for the identification of depression and sexual dysfunction; (ii) gynecological examination; (iii) biopsies for pathological examination. Prospective pts are seen by the gynecologists up to 3 years post-transplant. Clinical findings of genital cGvHD is a dry, thin, painful mucosa, inflammation, lichenoid bands, vulvar synechiae and vaginal stenosis. Systemic and local estrogens do not resolve those signs. Histology findings were those of chronic inflammation, lichenoid reaction and fibrosis. Results: Fifty-eight consecutive pts with 1-10 yrs of follow-up were approached and so far 30 women have been examined. Median age was 46 (26-71) yrs) and follow-up post-transplant was 5.5 (1-10) yrs. Fifteen pts (50%) had a clinically diagnostic cGvHD, and using pathology examination (n=13) the diagnosis was confirmed (n=6) or suspected (n=4). Another 12 of the 30 pts had signs suggestive of cGvHD and histology was confirmatory in one and suspect in 6 cases. Two of 30 pts had vulvar synechiae and 10 (33%) had vaginal adhesions. In the prospective study, 19 pts have been included and the follow-up is now 12 (3-24) months. After 12 months of follow-up (n=14) clinically cGvHD had been diagnosed or suspected in 8 cases. Local corticosteroid or takrolimus cream was commenced in 4/19 cases. Female Sexual Distress Scale intervention performed as sceduled revealed sexual dysfunction at at least once in 23 (46% of all) and Beck Depression Inventory indicated depression 18 (37%). Conclusions: Genital signs and symptoms, including vaginal stenosis, are common features of cGvHD and are associated with sexual dysfunction and depression. Gynecological surveillance and early intervention may reduce the risk of severe sequelae after genital cGvHD.

Langerhans cell chimerism early after T-cell depleted allogeneic haematopoietic stem cell transplantation K. Schneiker*, T. Schmitt, A. Konur, J. Hemmerling, K. Bender, E. von Stebut, A. Hadian, K. Kolbe, C. Huber, W. Herr, R.G. Meyer University Clinic Mainz (Mainz, DE) Skin is the most frequently affected organ in acute graft versus host disease (GVHD). Data from murine studies suggest that the interaction of residing host epidermal Langerhans cells (LC) and donor T cells is crucial for the initiation of acute GVHD. In an ongoing clinical protocol applying alemtuzumab-based T cell depleted (TCD) allogeneic stem cell transplantation (SCT) we observed acute skin GVHD occurring very early after transplantation despite of low T cell counts in the peripheral blood (Meyer, Blood 2007; 109:374) . We therefore intended to analyze the LC chimerism in these patients. Up to now, LC-chimerism analysis in humans has been performed by the detection of the sex-chromosomes restricting it to sex-mismatched donor / recipient pairs. In our patient-population, this would limit the analysis to less than 1/3 of the patients. Consequently, we aimed at a method to isolate LC from small skin samples with high purity for a sensitive STR-based chimerism analysis of general applicability. Epidermal skin layers were prepared from 6 mm punch biopsies by digestion with dispase I. They were further split and used for both immunofluorescent staining and digestion with trypsin to generate a single cell suspension and CD1a/MHC-class II-positive LC were subsequently sorted by flow cytometry. The density of LC on day + 20 after HSCT was much lower compared to before transplantation or to that of healthy individuals. But still, LC could be purified in all 10 analyzed patients. The isolated LC numbers ranged from 10 to >1000. We confirmed that the purity was exceeding 93.5% in 5 patients in a FACS-reanalysis, thereby reproducing the findings with skin of healthy individuals. Applying two alternative STR-based protocols, we obtained reliable results for LC chimerism in 8 of 10 patients and could detect signals with as few as 35 isolated cells. In 2 patients, the majority of isolated LC was of donor origin whereas the other 6 patients had predominantly host LC. After day +50 post HSCT, 2 further patients showed only a few remaining LC of host origin. In summary, we have established a sensitive method that enables the chimerism analysis on highly purified LC independent of sex-mismatched donor / recipient pairs. Our results on a few patients' samples cannot yet be related to clinical events. But the method allows the investigation of LC´s chimerism and potentially of other tissue-resident antigen presenting cells to study their impact on GVHD in humans. We and others have data on the activation of coagulation and fibrinolysis during and after allogeneic stem cell transplantation (SCT), which may have pathogenic implications in the subsequent recovery. We characterised the outcome including graft versus host disease (GvHD) in association with adaptive mechanisms of anti/coagulation after allogeneic SCT in 30 patients with a hematological malignancy. They were given myeloablative conditioning with cyclophosphamide and total body irradiation. 19 patients received the transplant from a sibling and 11 from an unrelated donor. GvHD prophylaxis consisted of cyclosporine were "unresponsive, non-progressive". The prognostic features, including cytogenetics, were similar in both groups. 70% of the patients responded to the first HDT (CR/nCR 8%, PR 48%, MR 13%). 37 patients were given a second transplant (26 "auto", 11 "allo"). 41% who received a second "auto" up-graded their response (CR 9%, PR 14%, MR 18%) while 42% who underwent "allo-RIC" increased their response (CR 28%, PR 14%). Median survival of the whole series was 3 years. Patients progressing while on therapy had a shorter survival than the "no-change" group (median 2 yrs vs not reached, p=0.00002). Finally, the 50 "non-responsive, nonprogressors" patients had similar survival than the 716 with chemosensitive disease intensified with HDT. Conclusions: 1) HDT in patients with primary refractory MM results in a low CR rate, 2) patients progressing while on initial therapy have a short survival despite the intensive approach and 3) patients with "non-responding, non-progressive" disease have similar survival than chemosensitive patients. Whether this good outcome is due to the impact of HDT or to the natural history of a more indolent disease remains to be further investigated. Background: Primary amyloidosis (AL) responds poorly to conventional therapy and has poor prognosis. Autologous stem cell transplantation (ASCT) results in a significant response rate although the procedure is hampered by a high transplant-related morbidity and mortality. Aim. To analyze the outcome of a series of 34 patients with AL who underwent ASCT at a single institution during a 10years period. Patients and methods: Thirty four patients (16 M, 18 F; median age 54 years, range: 33-66) who received an ASCT between November 1997 and September 2007 were included. Fourteen patients had received previous therapy and 20 were newly diagnosed. In 71% of the patients the light chain was of lambda type. The median number of involved organs was 2 (range, 1-4) including kidney (26 patients), heart (19), liver (11) , peripheral nerve (9), autonomic system (5) and gastrointestinal tract (5). Thirty-eight percent of the patients had more than 2 organs involved. All patients were mobilized with G-CSF alone and the intensive regimen consisted of MEL-200 in 23 patients and MEL-140 in 11 patients. The median interval between diagnosis and ASCT was 9 months. Results: The overall transplant-related mortality (TRM) was 27%. There was a trend towards a higher TRM in patients transplanted from November 97 -April 02 (n=14, TRM: 42%) as compared to those transplanted from June 02 -Sept 07 (n= 20, TRM 15%) (p= 0.07). In the 26 patients who underwent ASCT before Nov. 2006 (minimum follow-up of 1 year) the response and survival on an intent-to-treat basis were as follow: CR (28%), PR (12%), no response (26%), early death (34%).Organ response was observed in 11 patients (42%). The median response duration and the overall survival were 50 and 59 months, respectively. Favorable prognostic features for survival were: cardiac septum <14 mm (p= 0.03), normal beta 2m (p=0.04) and normal NT-proBNP (p= 0.006). Updated results on the overall series of 34 patients will be presented at the meeting. Conclusion: 1) ASCT in AL results in significant CR rate with prolonged response duration an overall survival and 2) although TRM is high, there is a trend towards a lower TRM over the years this most likely reflecting both a better patient´s selection and general management.

Outpatient-based peripheral blood stem cell transplantation for patients with multiple myeloma A. Ghavamzadeh, M. Khani*, A. Karimi, K. Alimoghadam, A. Manokian, R. Maheri, M. Asadi, F. Afshar, A. Shamshiry Hematology, Oncology and BMT Resarch Center (Tehran, IR) Intrduction: The aim of this study was to explore the feasibility and safety and cost-benefit of performing ASCT on an outpatient basis. Material and Methods:Total of 86 patients affected by MM and in complete remission (CR) or partial remission (PR) were selected to receive ASCT on an out-patient or in-patient basis .

In the in-patient group 31, 12 patients received 200mg/m² and 140mg/m² melphalan as conditioning regiment respectively . In out-patient group 12 patients received 140mg/m2 and 30 patients recived 200mg/m 2 melphalan. In out-patient group all the patients were programmed to go home the day after ASCT and to be rehospitalized in the case of febrile neutropenia or other sever toxicities. we used caregiver, general physician, staff nurse as an out-patient and visit team and also unequipped routine house of the patients during neutropenia. Results :Median ages were 50±7.5 years, median hospital stay were 28, 6.5 days in in-patient and out-patient respectively. There were not significant difference between these groups in aphresis days ,granulocyte colony stimulating factor(Gcsf) requirement for mobilization and mononoclear cell (MNC) or Cluster of differentiation(CD)34 + cell parameters (p<0.1), but statistically significant reduction in total cost and hospitalization (p<0.017).There were also significant reduction (p<0.001) in parentral antibiotic,blood product requirement and need for total parentral nutrition. Conclusion :Many different authores have explored the feasibility of autografting patient on an outpatient basis. The ease of administration of HDM as well as the lack of excessive extramedullary toxicity, including nausea and vomiting renders patients with MM more suitable for outpatient management, In the present study, we describe an outpatient program based on management of the patient in his/her house during aplastic phase . Our results clearly indicate that such a proceder is feasible and safe in a patient population with a median age of 55 years. With an accessible caregiver, the most frequent cause of readmission in other study were febrile neutropenia and sever mucositis need TPN. IN particular, it is worth nothing that transplant cost, requirement for blood product and hospital stay decreased significantly in this method of transplantation.

Treatment of multiple myeloma with sequential autologous and low dose total body irradiation based allogeneic unrelated SCT C. Pfrepper*, T. Lange, R. Krahl, M. Cross, H.-K. Al Ali, W. Pönisch, N. Basara, D. Niederwieser University Hospital of Leipzig (Leipzig, DE) Purpose: The limited availability of related donors poses an important limitation in the treatment of multiple myeloma (MM). The use of unrelated donors would alleviate this problem. Furthermore, outcome after unrelated transplants following reduced intensity conditioning (RIC) has been shown to be superior in other diseases including CLL and AML, due to increased graft-versus malignancy reaction. In this study, the feasibility of autologous followed by allogeneic unrelated RIC SCT was tested and the results compared to autologous followed by related allogeneic RIC SCT and allogeneic RIC SCT after relapsing following autologous SCT. Patients and Methods: We report the outcome of 39 MM patients with a median age of 55 (range 33-64) years treated with high dose chemotherapy (Melphalan 200mg/m² on day -3) followed by autologous SCT. Subsequent RIC-SCT was performed using Fludarabine (30mg/m²/d from days -3 to -1) and TBI (2 Gy on day 0) followed by cyclosporine (6.25 mg/kg twice daily from day -1) and mycophenolate mofetil (15 mg/kg daily from day 0). Eleven patients received an unrelated graft (group 1) and 18 a related graft (group 2) within median 5 (range 2-8) months from the autologous HCT. Ten further patients (group 3) underwent either related or unrelated SCT 13 (range 2-42) months after insufficient response following autologous transplantation and additional chemotherapy. Results: Durable engraftment was obtained in 97% of all patients. Overall survival (OS) at 2 years was 62±15%, 82±10% and 67±16% (p=0.089) in groups 1, 2 and 3, respectively at a median follow up of 23 (range 2-66) months. Nine patients (23%) were in CR, 15 patients (38%) in PR, 3 patients (8%) had progressive disease after allogeneic SCT. However; progression free survival (PFS) was 55±15%, 20±10% and 11±10% in the three groups, respectively (p=0.014). Non relapse mortality (TRM) was not different between the three groups, but relapse incidence was lowest in patients after unrelated transplantation. Conclusion: We conclude that autologous SCT followed by low dose TBI based SCT from matched unrelated donors provides rapid and sustained engraftment for MM patients comparable to that of related donors. Our limited data suggest a tendency to higher PFS following unrelated compared to related SCT. These results provide the basis for a phase III study designed to compare auto-allo unrelated to auto-auto SCT. Background: A significant proportion of patients with multiple myeloma have a long-lasting response after autologous stemcell transplantation (ASCT). However, others relapse relatively quickly. The aim of this study was to determine if the presence of monoclonal plasma cells (MPC), detected by flow cytometry, in the apheresis product, and in pre-and posttransplant bone marrow samples predicts a shorter time to progression (TTP). Patients and Methods: We included patients diagnosed with multiple myeloma and treated with high dose therapy followed by ASCT between November 1, 1998 and February 1, 2007. In all casas, flow cytometric analysis was performed in the apheresis products and in bone marrow samples, both prior and after ASCT. The variables evaluated as possible prognostic factors were: age, sex, type of monoclonal component, Durie-Salmon stage, ISS stage, presence of MPC (CD138+/CD38+/ CD19-/CD45-or CD45dim) in apheresis products, and in bone marrow samples (30 days before, and 100 days after ASCT), and accomplishment of a complete response prior or after ASCT. Results: 55 patients were included: 21 male, 34 female. Median age at ASCT was 58 years (range, 30-69 years). Twenty-nine patients had progressed. On univariate analysis, age .14); p<0.023], ISS stage [OR 7.61 (2.14-27.14); p=0.08], MPC in apheresis products (p=0.005), MPC in pretransplantation bone marrow (p=0.007), MPC in postransplantation bone marrow (p=0.06), and the achievement of complete response prior and after ASCT (p=0.01) had a negative impact on time to progression. On multivariate analysis, the presence of MPC in apheresis [OR 3.1 (1.24-2.65), p=0.01] and the presence of MPC in pretransplantation bone marrow (OR 14.2, p=0 .012) were identified as independent predictors of a shorter TTP.

Conclusions: The presence of MPC in apheresis products and in pretransplantation bone marrow were identified as independent predictors of shorter TTP. Both parameters can identify a group of patients with multiple myeloma which could benefit from most aggresive conditioning regimens or additional maintenance therapies.

Long-term follow-up of patients with systemic AL amyloidosis treated with high-dose melphalan and autologous stem cell transplantation after induction and mobilisation chemotherapy S. O. Schonland*, T. Bochtler, M. Hansberg, A. Mangatter, J.B. Perz, A.D. Ho, H. Goldschmidt, U. Hegenbart University of Heidelberg (Heidelberg, DE) Introduction: Longterm survival in AL amyloidosis (AL) patients (pts) has been shown recently after high-dose melphalan (HDM) (Sanchorawala, Blood 2007) in 21% of the patients. However, no advantage of HDM compared to conventional chemotherapy has been observed in a randomized multicenter trial (Jaccard, NEJM 2007) . This was probably due to high transplantation related mortality (TRM) and a substantially longer time to the start of chemotherapy in the HDM group. Whether treatment intensification using induction and mobilization chemotherapy prior HDM can improve results has not yet been definitively shown. Methods: We have updated 23 pts with AL (Perz, BJH 2004) who received vincristine/adriamycine/dexamethasone as induction and additional mobilization chemotherapy (mostly cyclophosphamide based) prior to HDM in our center from 1998 until 2004. Inclusion criteria were age < 70 years, NYHA stage < III and a WHO performance status < 3. Median age was 57 years; median number of involved organs was 3. Fifteen and 13 pts had symptomatic renal or cardiac involvement, respectively (on dialysis at HDM, n=4). Results: The median overall survival (OS) has not been reached at a median follow up of 66 months post HDM (figure). TRM was 9%. A hematological response (HR) to induction and mobilization chemotherapy occurred in 47% (14% of pts achieved a complete remission (CR)). HR and CR rate increased to 84% and 67% after HDM, respectively. Median time to CR from HDM was 4 months. Two patients had already an organ response (OR) after induction and mobilization chemotherapy. OR was observed in 68% and the median time to OR was 11 months post HDM (range 0-40 months). Pts with CR have an estimated OS of 90% after 6 years; for pts not reaching CR the median OS is 48 months (p=0.002). Hematological relapse or progression occurred in about half of the patients; however 6 pts (26%) are still in first CR with a sustained OR and a median follow-up of almost 7 years. Conclusion: We observed a very high CR and OR rate using this intensive treatment approach. Patients with CR after HDM have an excellent survival. Therefore, the role of induction therapy in AL, e.g. with bortezomib or lenalidomide, should be further investigated in randomized trials in experienced centers. The main goal of treatment in AL remains achievement and maintenance of a CR of the underlying plasma cell disorder. The results of reduced-intensity conditioning allogeneic stem cell transplantation (RIC allo-SCT) for multiple myeloma (MM) are still under considerable debate. While EBMT data did not support the universal use of RIC for MM allografts, the Italian randomized multicenter study suggested that in newly diagnosed myeloma, survival in recipients of a hematopoietic stem-cell autograft followed by RIC allo-SCT from an HLAidentical sibling is superior to that in recipients of tandem stem-cell autografts. The aim of this multicenter retrospective national study was to identify prognostic factors for outcome of high-risk patients with MM after allo-SCT prepared by RIC. Data from 219 patients (median age 52 years, range 27-66), who received grafts from a sibling (n=197) or unrelated donor (n=22) were analyzed. At time of transplant, only 37 patients (17%) received RIC allo-SCT in CR or VGPR, while 134 patients (61%) were transplanted in PR. 48 patients were transplanted either in stable disease (n=15) or were in refractory/progressive disease (n=33). All patients have received at least one autologous transplant prior to RIC allo-SCT. The graft source was PBSCs in the majority of patients (n=183). 21% of the patients received the Seattle Fludarabine and low dose TBI RIC regimen, while 53% of patients received Fludarabine, Busulfan and ATG. 32 patients (15%) died of transplant-related complications. The incidences of grade 2-4 acute GVHD and extensive chronic GVHD were 37% and 20% respectively. At 3 years, overall and progression free survivals (OS, PFS) were 41% (95%CI, 34-49) and 19% (95%CI, 14-27) respectively. Disease status (CR, PR, SD vs. progressive) was significantly associated with overall survival (P=0.0002; Fig. below) . In multivariate analysis, disease status at time of RIC allo-SCT, was the strongest parameter associated with an improved OS and PFS (P=0.005 and P=0.004 respectively). Despite its obvious caveats, the relatively low TRM observed in this series, suggest that there is still space to investigate RIC allo-SCT for MM. However, RIC allo-SCT appears to result in a durable response only if it is applied early in the disease history, especially when patients are still chemosensitive. Since the latter results are also expected to be further improved with the systematic and early use of maintenance therapies (Bortezomib and/or Lenalidomide) after RIC allo-SCT, randomized or quasirandomized prospective studies are still warranted. The role of Reduced-intensity conditioning (RIC) allogeneic stem cell transplantation (allo-SCT) for adult patients with acute lymphoblastic leukaemia (ALL) is still under debate. ALL encompasses a group of chemosensitive diseases, raising concerns that significant reduction of the intensity of the preparative regimen may have a negative impact on leukemic control. In this multicenter retrospective study, the outcome of 601 adult (age at transplant >45 y.) patients with ALL who underwent transplantation in complete remission (CR) with an HLA-identical sibling donor, were analyzed according to 2 types of conditioning: RIC in 97 patients, and standard myeloablative conditioning (MAC) (or high-dose) in 504 patients. Both groups were comparable in terms of gender, CR status (CR1 and CR2), interval from diagnosis to allo-SCT, and R/D CMV serostatus. Patients in the RIC group were older (median 56 y. vs. 50y in the MAC group;P<0.0001). Most of the patients in the MAC group received high dose TBI (80%), while the majority of the RIC regimens included either low-dose TBI or were ATG+chemotherapy-based regimens. The majority of patients (88%) from the RIC group received PBSCs. In the MAC group, the stem cell source consisted of bone marrow in 42% of patients. With a median follow-up of 13 months (range, , the incidences of grade II-IV and grade III-IV acute GVHD were: 35%, 14%, and 28%, 10% in the MAC and RIC groups respectively (P=NS). The cumulative incidence of nonrelapse mortality at 2 years (NRM) was 32% (MAC) vs. 22% (RIC) (P=0.04). The cumulative incidence of relapse at 2 years was 30% (MAC) vs. 42% (RIC) (P=0.0007). However, the latter differences did not translate into any significant difference in term of leukemia-free survival (LFS) at 2 years: 38% (MAC) vs. 37% (RIC) (P=0.42). In multivariate analysis for LFS, the status at transplant was the only factor associated with an improved LFS (p<0.0001, RR=0.55, 95%CI, 0.42-0.72). The results of this study suggest that RIC regimens may reduce NRM rate after allo-SCT for adult ALL when compared to standard MAC regimens, but with a higher risk of disease relapse and no impact on LFS. The latter represent promising findings, since patients who received RIC are likely to have serious comorbidities, which led the transplantation center to choose RIC, and surely most of these patients would not have received a standard allo-SCT in most institutions. Therefore, RIC allo-SCT for adult ALL (>45 y.) may represent a valid therapeutic option when a conventional standard conditioning is not possible, warranting further prospective investigations.

Reduced-intensity conditioning allogeneic stem cell transplantation for patients with acute myeloid leukaemia: long-term results of a "donor" versus "no donor" comparison M. Mohty *, H. De Lavallade, J. El-Cheikh, P. Ladaique, C. Faucher, S. Fürst, N. Vey, D. Coso, A.M. Stoppa, J.A. Gastaut, C. Chabannon, D. Blaise Institut Paoli-Calmettes (Marseille, FR) The issue of possible higher relapse rates after reducedintensity conditioning (RIC) allogeneic stem cell transplantation (allo-SCT) is still under considerable debate. This report describes the updated long term results (initial publication, Leukemia 2005) of 95 consecutive acute myeloid leukaemia (AML) patients, diagnosed between 1999 and 2003 in a single centre. Using a genetic randomization through a "donor" versus "no donor" comparison, our aim was to assess the true benefit of RIC-allo-SCT for adult AML. In this series, 35 patients (37%; "donor" group) had an "identified" HLAidentical sibling donor, while the remaining 60 patients had no HLA-matched related donor ("no donor" group). As per institutional policy, HLA-MUD were not considered during the study period. No significant differences in patients or AML features were found between the two groups. In the "donor" group, 25 patients (71%; median age, 51 (range, 26-60)) could actually proceed to the RIC-allo-SCT. The current median follow-up is 60 months. In an "intention-to-treat" analysis, the KM estimate of leukemia-free survival (LFS) was significantly higher in the "donor" group as compared to the "no donor" group (P=0.003; 60% versus 23% at 7 years). When restricting the analysis to patients who could actually receive the RIC-allo-SCT (median follow-up, 40 months from time of allo-SCT), the difference in LFS was also significant between this group of 25 patients ("transplant" group) and the remaining 70 patients ("no transplant" group; P=0.0002; 72% versus 24% at 7 years). No major toxicities were encountered during RIC administration (fludarabine, busulfan and ATG), and only 3 patients died from toxicity, for a cumulative incidence of TRM of 12% (95%CI, 3-32%) at last follow-up. This relatively low TRM translated towards a significantly higher overall survival (OS) in the "transplant" group as compared to the "no transplant" group (P=0.0003). In the "intention-to-treat" analysis, OS was still significantly higher in the "donor" group as compared to the "no donor" group (P=0.003; Fig. below) . After controlling for all relevant factors, in the multivariate analysis, only actual performance of RIC-allo-SCT (P=0.0005; RR=4.1; 95%CI, 1.8-9.1), was significantly predictive of an improved LFS. Based on these long term results, we conclude that if a matched related donor is identified, RIC-allo-SCT should be proposed since it represents a valid and potentially curative option for AML patients not eligible for standard myeloablative allo-SCT.

Graft failure after reduced-intensity conditioning B. Hertenstein*, E. Dammann, R. Brand, A. van Biezen, D. Niederwieser, T. de Witte, T 

Reduced intensity conditioning (RIC) regimens use doses of chemotherapy and/or radiotherapy which per se do not eradicate the malignant cells but which should be nevertheless sufficient to allow sustained engraftment. The incidence of graft failure might therefore be higher in RIC and may be dependent on the regimen used. To evaluate the incidence of graft failure and the potential risk factors in patients transplanted after RIC, data files of such transplantations were selected from the CLWP data base. In all records chemotherapeutic drugs and dosages as well as TBI dosages were individually checked and only records meeting the EBMT definition of RIC were included. The analysis was further restricted to patients surviving > 28 days. 1720 RIC transplants were identified. A chemotherapeutic RIC regimen was used in 1066 patients (Flu/Bu 571, Flu/Mel 296, Flu/Cy 60), TBI only in 97 and a combination of chemotherapy and TBI in 557 patients. Stable engraftment occurred in 1640 patients (95.3%), no engraftment in 45 patients (2.6%) and a graft loss in 35 patients (2.0%). Survival for patients with graft failure was 76% at 100 days and 38% at one year. 29 patients received a second transplant and survival at one year was 60% for these patients. Graft failure was lowest in patients with myeloma (1.6%) and highest in patients with MDS/MPS (8.7%). With the use of bone marrow as graft source graft failure was much higher than with peripheral blood stem cells (11.8% vs. 3.7%) . Graft failure occurred in 3.4% of transplants from HLA-identical siblings, in 6.6% from matched unrelated donors and in 12.3% from mismatched unrelated donors. There was no difference between the different conditioning regimens, neither between chemotherapy alone, TBI or combined regimens (3.4% vs. 2.3% vs. 4.4%) nor between the major chemotherapy regimens themselves (Flu/Bu 1.9%, Flu/Mel 2.6%, Flu/Cy 4.3% for HLA-id siblings). ABO match and donor-recipient sex mismatch had no impact on graft failure. In multivariate analysis underlying disease and graft source were significantly correlated with graft failure. In summary the incidence of graft failure after RIC is low. MDS/MPS as underlying disease and the use of bone marrow as graft but not the type of the conditioning regimen were risk factors for graft failure. (2), H. Esperou (2) , M. Attal (2) , N. Milpied (2) , B. Lioure (2) , P. Bordigoni (2) , I. Yackoub-Agha (2) , J. Bourhis (2) , B. Rio (2) , E. Deconninck (2) , M. Renaud (2) , N. Raus (1) , D. Blaise (2) (1)Hôpital Edouard Herriot (Lyon, FR); (2)Société Française de Greffe de Moelle et de thérapie cellulaire (Saint-Denis, FR) This report updates a retrospective study from SFGM-TC registry concerning 1108 patients who underwent allogeneic hematopoeitic stem cell transplantation (HSCT) after reduced intensity conditioning (RIC) from HLA identical siblings (84%) and unrelated donors (16%) for hematological malignancies. At time of conditioning, 442 patients were in CR, 337 in PR, 107 in stable disease (SD) and 222 in progressive disease (PD). As conditioning, 255 patients received fludarabine and TBI (2 grays), 465 patients fludarabine, busulfan and ATG and 388 patients other regimens. After transplant, 336 patients (30%) developed an acute GVHD grade II (grade II: 178, III: 80 and IV: 78). A chronic GVHD was present in 388 patients (35%) (185 limited and 203 extensive). With a median followup of 30 months, the 3 and 5-year probability of overall survival (OS) were 43.5% (40-47) and 32%(29-35) respectively and the 3 and 5-year probability of event-free survival (EFS) were 35%(31-39) and 28% (24.5-31) respectively. The TRM at 1 year, 2 years and 3 years was 15% (13-17), 18% (15.5-21) and 20% (17-23). A mixture model, gfcure with Splus statistical package determined the percentages of long-term survivors and its adequacy was verified graphically. The probability to be a long-survivor was 24% (17.5-32.5) (Fig.1 ) and to be a long event-free survivor was 23% (19-28) (Fig. 2) . The multivariate analysis has tested recipient and donor age, disease status pre-transplant, number of transplants before RICT, HSC source, sex matching, HLA matching, CMV status and ABO compatibility. The only factor which had a significant impact on long-term survival after RICT was the disease status just prior conditioning: PR versus CR: HR: 3.63 [1.14-9.18] p<0.001 and PD versus CR: HR: 4.35 [2.22-8.51] p<0.0001. In conclusion, these updated data demonstrate that allogeneic HSCT after RIC was able to possibly cure 23% of patients with haematological malignancies and the most important factor to take into account remains to be in CR pre-transplant.

The haematopoietic cell transplantation-specific comorbidity index predicts survival and non-relapse mortality in lymphoma and myeloma patients receiving a RIC allograft L. Farina* (1), B. Bruno (2) The allogeneic hematopoietic cell transplantation-specific comorbidity index (HCT-CI) has been recently developed to identify patients at high risk of morbidity and mortality after an allogeneic stem cell transplant (alloSCT). Reduced-intensity conditioning (RIC) regimens have decreased non-relapse mortality (NRM) in heavily pre-treated patients. We performed a retrospective study to assess whether comorbidities, according to HCT-CI, may influence the outcome of lymphoma and multiple myeloma patients undergoing a RIC alloSCT. Between 2000 and 2007, 197 patients received a RIC alloSCT from a HLA identical sibling (n=123) or unrelated (n=74) donor in three Italian Transplant Units. Median age at transplant was 52 years (range, 17-69) and 41% of the patients were 55 or older. Diseases included non Hodgkin's lymphoma (n=103), multiple myeloma (n=68) and Hodgkin's lymphoma (n=26). Median number of previous treatments was 3 (range, 0-8). Disease risk according to Kahl et al. (Blood 2007) was low, intermediate and high in 29%, 42% and 29% of patients respectively. Patients with HCT-CI of 0, 1-2 and ≥ 3 were 64 (32%), 61 (31%) and 72 (37%), respectively. Variables included in multivariate analysis were age (< 55 or ≥ 55), disease risk (low, intermediate and high), number of previous lines of therapy (≤ 2 and >2), HCT-CI (0, 1-2 and ≥ 3), Karnofsky Perfomance Status (PS, >80% and ≤ 80%). Oneyear OS and PFS were 87%, 60%, 60% and 91%, 75%, 70% in patients with HCT-CI of 0, 1-2 and ≥ 3 respectively. Cumulative incidence of NRM was 6%, 22%, 25% at 1 year and 6%, 24% and 27% at 2 years, whereas relapse mortality was 5%, 20%, 17% at 1 year and 7%, 27% and 23% at 2 years. By multivariate analysis only Karnofsky PS (p=0.00051) and HCT-CI (p=0.0038) were correlated with OS, whereas PFS was influenced by disease risk category (p=0.013), number of previous therapies (p=0.038), and both Karnofsky PS (p=0.031) and HCT-CI (p=0.0009). Interestingly HCT-CI was the only significant factor that could predict NRM (p=0.034) while Karnofsky PS failed to show a significant correlation (p=0.1). Of note, age (≥ 55) was not statistically significant either in OS, or PFS or NRM. Although the data need to be confirmed in prospective trials, these results showed that HCT-CI may be a useful tool to predict OS, NRM and also PFS after RIC alloSCT in lymphoma and myeloma patients. In the clinical setting HCT-CI and not age should be used for patient risk assessment. Venous thromboembolism (TE) occurs as a consequence of genetic and environmental factors. Important genetic risk factors are deficiencies of natural anticoagulants, antithrombin-III (AT-III), protein C (PC) and PS, and genetic mutations of factor V leiden (FVL) and prothrombin (PTH A20210). Thrombotic complications after HCT are usually catheter-related thrombosis (CRT), pulmonary TE (PTE) and deep vein thrombosis (DVT). Our aim in this study was to evaluate the effects of the deficiencies of ATIII, PC and PS, and the gene mutations of FVL and PTH on the incidence of the development of TE complications and liver sinusoidal obstruction syndrome (SOS) at the early or late period post-HCT. In our center, pre-transplant work-up includes routine thrombophilia tests. Between Apr 1999-Jan 2007 260 patients (M/F: 145/115, median age: 34 years) admitted to our transplantation center were retrospectively analyzed for the relation of the presence of thrombophilia and the frequencies of the occurrence of a TE complications and liver SOS. All but 6 patients (n=254) had an HLA identical sibling donor. The ratios of the detection of the plasma activation level below 50 % for PC, PS and AT-III were 5.8 % (12/206), 15.7 % (32/204) and 25.4 % (66/260), respectively. Gene mutations were studied in 198 patients prior to transplant and the frequencies of gene mutations were detected in 11.6 % patients (n=23) either FVL (n=14) or PTH (n=9) gene. None of the patients had both mutations, FVL and PTH. At the peri-or posttransplantation period we observed venous TE in 24 patients (17 CRT, 2 PE and 5 DVT), liver SOS in 23 patients and myocard infarction at the early period in only one patient. In 4 out of 12 patients with low PC activity CRT occurred, and liver SOS was observed in only one patient. In 6 out of 24 patients with genetic mutation developed a TE complication, 5 CRT and 1 PTE. CRT occurred in 4 of 14 patients with FVL mutation, while 1 CRT and 1 PTE among 9 patients with PTH mutation were observed. Liver SOS was seen in only one patient with PTH heterozygote positive. We found the presence of low PC level or genetic mutation increased the frequency of the development of TE (OR: 6.9 and 3.7, respectively), but no effect on the liver SOS. In conclusion, the use of thromboprophylaxis at peri-transplant period in patients with genetic mutations is still a controversial topic; which should be elucidated with controlled studies. Iron overload (IO) is an adverse prognostic factor in patients who undergo allogeneic haematopoietic stem cell transplantation (HSCT) for thalassemia and appears to play a similar role in patients with other hematological disorders. Estimation of IO is primarily based on serum ferritin, however many confounding factors particularly in HSCT recipients may result in frequent ferritin overestimation. Aim of the study was to quantify IO by uperconducting quantum interference device (SQUID) after HSCT and evaluate the impact on infections and GVHD; additionally the feasibility of iron-depletion has been investigated In this pilot study we evaluated IO in 89 consecutive adult patients who received HSCT from a matched sibling (n=66) or a matched unrelated donor (n=23). Primary diagnosis included AML/MDS in 54% of cases. Assessment of IO after HSCT included serum ferritin and in those with hyperferritinemia (> 1000 ng/ml), liver iron concentration (LIC) was evaluated by SQUID magnetic susceptometry. IO assessment was performed in patients in remission at a median time of 698 days after HSCT (range 48-5239 days). Median serum ferritin was 821 ng/ml (range 18-11110).Thirtynine of 89 (44%) patients had serum ferritin level >1000 ng/ml; LIC (microgFe/g liver wet-weight) evaluated by SQUID was available for 35/39 patients with elevated ferritin values. Overall, 26 patients (29%) had moderate (LIC levels 1000-2000 microg/gww) to severe (LIC levels>2000 microg/gww) IO; median LIC values were 1419 microg/gww (range 1030-3253). Nine patients (10%) had normal LIC values (LIC<400) despite high ferritin. Patients with LIC>1000 received a median of 41 packed red blood cells before study evaluation (range 7-79). The rates of bacteremias, invasive fungal disease and chronic GVHD were higher among patients with LIC>1000 as compared to patients with LIC<1000 (24%, 11% and 41% vs 16%, 5% and 25% respectively; p=NS). Eighteen of the 26 patients with LIC levels > 1000 were treated by regular phlebotomy. For 15 of these the phlebotomy program is still ongoing, while 2 patients completed the program with ferritin normalization; one patient was switched to deferasirox due to poor tolerance. In 5 cases phlebotomy was considered contraindicated; 3 patients who relapsed did not receive any treatment.

Our preliminary data show that one third of HSCT recipients may present moderate/severe IO as assessed by SQUID, and iron-depletion resulted feasible in 73% of these subjects. A trend for higher rates of infectious complications and chronic GVHD was observed in patients with IO. Despite recent advances, mortality rates after allogeneic hematopoietic stem cell transplantation (HSCT) remains high and cannot be accurately predicted. Several reports from the Seattle group suggested the use of comorbidity indexes (CI) to provide valid and reliable scoring of pre-transplant comorbidities that predicted non-relapse mortality (NRM) and overall survival [Blood 2004 (CCI) , 2005 (HCT-CI) & Ann Int Med 2006 (PAM)). However, whether such indexes could be used by other groups, and if one index better predicts survival than another, is yet unknown. HCT-CI and PAM required grading according to pulmonary function tests (PFT), which were lacking for a majority of our patients. We thus designed a modified HCT-CI and a corrected PAM, without PFT. Survival curves were estimated using Kaplan-Meier method. Cumulative incidences (CI) of non-relapse mortality were analyzed, with relapse treated as competing event. The likelihood ratio statistics in proportional hazards models was computed for each index, with its associated P-value, as a measure of association between the comorbidity indexes collapsed into risk groups and the outcomes. Discriminative performance of comorbidity indexes was then evaluated by the c-index. Standard errors of c-indexes were computed from that of Somers' Dxy rank correlation coefficient. P-values for comparison of c-indexes were obtained using a nonparametric bootstrap procedure. We thus retrospectively studied 286 patients (169 male, 117 female) who received their first allogeneic HSCT at Saint Louis Hospital between April 2004 and December 2006. The median age was 31 years (4-64), 149 patients (52%) were transplanted from a matched related donor and 248 patients received an HSCT for intermediate or high-risk diseases (86%). Using CCI, 25% of our patients had indexes of 1 or more; median reduced HCT-CI was 1; and median corrected PAM score was 24. The discriminative properties of the 3 CI were rather low in our patient population with c-index of 0.515, 0.499 and 0.582 for the CCI, reduced HCT-CI and corrected PAM indexes, respectively. Comparison of patients-and transplant-characteristics between our and Seattle group's cohorts however revealed significant differences with more children, more cord blood HSCT and HSCT for Fanconi Anemia in Saint Louis. Finally, direct comparison of scoring items and multivariate analysis revealed that age, unmatched related donor and hepatic disease were associated with NRM in our cohort. Primary graft failure occurs in 10-20% of unrelated cord blood transplantation (UCBT) patients. Its occurrence is associated with increased risk of death due to relapse (when associated with autologous recovery) or complications of pancytopenia. Salvage transplant is possible in patients with absence of neutrophil recovery (ANR): clinical tools are required to promptly identify these patients at risk. We conducted this retrospective study to identify risk factors for ANR. From 1996 to 2007, 98 consecutive children underwent a first UCBT in our center. ANR was defined as absolute neutrophil count ≤ 100/microL in patients alive without disease on day +42. Patients for whom a salvage transplant conditioning regimen was initiated before day +42 were excluded from this analysis (n=4, 2 of them are late survivors). Ninety one (91) children were eligible for analysis. Median time for neutrophil recovery (≥ 500/microL) was 27 days (range, 0 to 51). ANR occurred in 7/91 (7.7%) and was confirmed by bone marrow biopsy. Of these 7 patients, 2 underwent a 2nd transplant, one of which succeeded. Five (86%) died between day +48 and +72 of transplant complications that occurred at a median of 38 days (range 29-70) of their first transplant. In 4 of them, complications were related to persistent neutropenia and/or lymphopenia. Our current strategy of waiting until day +42 for a 2nd transplant decision carries a significant risk of mortality in patients with ANR. When analyzed for demographics and graft characteristics, the group with ANR did not differ significantly from the group with neutrophil recovery. The specificity and sensitivity for ANR prediction of White Blood Cell (WBC) < 200 /microL at different time points is as shown in table 1. WBC < 200 /microL on day +28 and day +35 carries a risk of day +42 ANR of 58% and 100%, respectively. The negative predictive value of WBC < 200 /microL on day +35 is 100%. Therefore, in order to reduce the risk of mortality associated with ANR, we now proceed to 2nd transplant evaluation for children with WBC < 200 /microL by day +28, and initiate the therapy on day +35 if WBC counts remains < 200 /microL. Purpose: The aim of this retrospective study was to determine whether pre-transplant iron overload assessed by serum ferritin level predict the non-relapse mortality (NRM) and overall survival (OS) rate after haematopoietic cell transplantation (HSCT). Patients and Methods: 283 patients with hematological disorders underwent myeloablative or reduced-intensity conditioning followed by the first HSCT from related or unrelated donors at Keio BMT Program from 1995 to 2005. 35 patients were excluded from this analysis because ferritin and CRP levels were not available or measured at the time of transplant. Demographic data and information on comorbidities was obtained from the Keio BMT database. Comorbidity was scored according to the HCT-specific comorbidity index (HCT-CI) proposed by the Seattle group without any modifications. Results: Median value of ferritin was 448 ng/mL(range 26-3010 ng/mL). The percentage of patients with ferritin 193 ng/mL did not vary significantly among diseases. In all but 7 patients CRP were within normal range at the time of measuring ferritin. There was a strong relationship between pretransplant ferritin and OS. The 5-year OS for patients was 76.2% in the first quartile (ferritin 0-193 ng/mL) (Q1), 58.6% in Q2 (194-448 ng/mL), 50.7% in Q3 (449-1060 ng/mL), and 43.6% in Q4 (more than 1061ng/mL) (P=0.001). 5-year OS was also significantly higher for those with ferritin level more than 1000 ng/mL compared with those with ferritin equal or less than 1000 ng/mL (61.8% vs 43.8% P=0.004). The 5-year NRM was also significantly associated with pretransplant serum ferritin level. The 5-year NRM was 12.2% for patients with ferritin more than 1000 ng/mL, and 28.8% for those with ferritin equal or less than 1000 ng/mL (P=0.004). The mean score of HCT-CI was 1.3 for those in Q1, 1.6 in Q2, 1.6 in Q3, and 2.5 in Q4. The mean score of HCT-CI was significantly higher in patients with hyperferritinmia (ferritin more than 1000 ng/mL) than those without (2.5 vs 1.5 P=0.001). Conclusion: These results suggested that pretransplant level of ferritin could be a marker of predicting NRM, OS after HCT and thus useful for patient counseling before HCT. Whether the addition of ferritin level into HCT-CI scoring system give a better prediction of posttransplant NRM and OS needs further investigation.

Increased incidence of renal impairment after radioimmunotherapy as part of the conditioning regimen before allogeneic stem cell transplantation T. Zenz* (1), R. Schlenk (1) Intensifying the conditioning regimen with radioimmunotherapy (RIT) is feasable for high risk leukemias and MDS patients. While additional radiation exposure by RIT is greatest in the bone marrow, significant exposure of the kidney occurs and an increased frequency of BMT nephropathy has been observed in patients receiving RIT using the Rhenium188 labelled anti-CD66 antibody. In order to obtain a precise picture of the incidence of renal impairment we compared two large cohorts receiving allogeneic stem cell transplantation (SCT) with or without the use of RIT from our center. Between 1998 and 2004, 261 patients with a median follow-up of 58 months received HLA-identical or mismatched allogeneic SCT. Of these 123 received an intensified conditioning regimen with a Rhenium188 (n=87) or Yttrium90 labelled (n=36) anti-CD66 antibody. The cohorts were similar with respect to sex, stem cell source and use of nephrotoxic medication. There was an increased proportion of patients with AML, ALL, and haploidentical donors in the group receiving RIT (p<0.01), while the conventional group had more patients with CML, TBI in the conditioning regimen, cyclosporine use and incidence of GvHD (p<0.01). The clinical characteristics were documented for 5 years post SCT. We defined relevant renal insufficiency as a creatinine level above 200 µmol/l and the failure to reach creatinine values below 150 to exclude intermittent drug toxicities. The incidence of renal failure was 13% in patients with conventional conditioning regimens and 24% in patients receiving RIT (p=0.02). The competing risk analysis showed a significant increase in the cumulative incidence of nephropathy (p=0.01). Even though the dosimetry showed decreased renal doses in patients with Y90 this did not lead to a decreased incidence of renal impairment. Other factors as TBI, cyclosporine use, donor source or GvHD did not influence the development of renal failure. In a multivariate model including RIT, CSA, donor source, GvHD and the combination of CSA and RIT, the combination of RIT and CSA was the only significant risk factor (p=0.01).

In conclusion, RIT lead to an increased frequency of continuous renal impairment. The combination of cyclosporine with RIT appeared detrimental while TBI (with renal shielding) did not lead to an increased incidence of renal failure.

Strategies to decrease the incidence of renal impairment may include nephroprotection with ACE inhibitors or alternative immunosuppression.

Prospective evaluation of oral mucositis in allogeneic stem cell transplant recipients receiving conventional myeloablative conditioning or reduced-toxicity conditioning with treosulfan and fludarabine H. Uotinen*, L. Volin, E. Juvonen, A. Nihtinen, T. Ruutu Helsinki University Central Hospital (Helsinki, FI) Oral mucositis (OM) is a frequent and often severe complication of allogeneic stem cell transplantation (SCT). The aim of this study was to evaluate the incidence and duration of severe (WHO Grade III-IV) and ulcerative (Grade II-IV) OM in patients receiving different conditioning regimens. Between Sept 2005 and Nov 2007, 131 patients were prospectively observed daily from the start of conditioning until hospital discharge. Conventional (convent.) myeloablative (MA) conditioning (Cy/TBI n=82, BuCy n=10) was given to 92 patients (34 AML, 21 ALL, 7 MDS, 7 CML, 7 MM, 6 MF, 10 others) and treosulfan (Treo)-based conditioning to 39 (13 MDS, 12 AML, 9 MM, 5 others) patients. Treo-based conditioning consisted of Treo 3x14 g/m² (n=28) or 3x12 g/m² (n=11) and fludararabine (Fld) 150 mg/m². Treo-Fld conditioning is regarded as MA at least at the Treo dose level of 3x14g/m² but with reduced non-hematological toxicity. Treobased conditioning was given to patients not eligible for convent. full intensity conditioning because of age, comorbidity or heavy preceding cytotoxic treatment. There were no significant differences in the OM results with the two Treo dose levels used. The median age of the patients was 48 (18-62) years in the convent. MA group and 57 (19-65) years in the Treo group. In the convent. MA group the donor was sibling in 40 cases and unrelated in 52 cases. In the Treo group the donor was sibling in 17 cases and unrelated in 22 cases. The incidence of severe OM was 40 % in the convent. MA group and 3 % in the Treo group (p< 0.005). Severe OM episodes had a mean duration of 4 (1-26) days. The incidence of ulcerative OM was 72 % in the convent. MA group and 33 % in the Treo group (p<0.005). The ulcerative OM episodes had a median duration of 6 (1-34) and 3 (range 1-8) days (p=0.003), respectively. OM reached maximum WHO grade on day 13 in the convent. MA group and on day 16 in the Treo group after the start of conditioning. The median duration of hospitalisation from the SCT was 24 (17-52) days in patients with ulcerative OM vs. 21 (range 17-43) days in patients without (p=0.110). In conclusion, conditioning with treosulfan and fludarabine caused significantly less severe and ulcerative mucositis than conventional MA regimens, and the duration of ulcerative mucositis was shorter. This supports the role of Treo-Fld conditioning for allogeneic transplantation especially in the treatment of patients with increased risk of toxic complications.

Patients receiving CBT had significantly slow neutrophil and platelet recovery in multivariate analysis. The incidences of acute and chronic GVHD were not significantly different. Unrelated BMT showed better TRM (25% versus 38% at 1 year, P<0.01), relapse (15% versus 26% at 3 years, P<0.01) and DFS (57% versus 29% at 3 years, P<0.01) results compared with CBT. Next, we analyzed 498 BMT (n=370) and CBT (n=128) recipients who have taken total body irradiation (TBI; >8Gy) containing myeloablative regimen and calcineurin inhibitors (cyclosporine or tacrolimus) plus methotraxate (MTX) for GVHD prophylaxis without history of prior transplants. In this subpopulation, multivariate analysis revealed no significant difference between BMT and CBT in DFS (57% and 51% at 3 years; hazard ratio (HR):1.34; 95% confidence interval (CI): 0.89-2.01; P=0.16). No statistically difference was also seen in TRM (25% at 1 year after BMT and 23% at 1 year after CBT; HR: 0.90; 95%CI: 0.58-1.39; P=0.64). However, the cumulative incidence of relapse was significantly lower in BMT than in CBT (16% and 21% at 3 years; HR: 1.90; 95%CI: 1.12-3.22; P=0.017). The current Japanese registration data in MDS showed overall results of unrelated BMT were better than those of unrelated CBT by competing risk regression models. These data also suggest that unrelated CBT could be safely and effectively used as same as unrelated BMT when adequate transplant procedures are selected.

comparative studies have shown that CB transplant is characterized by a lower risk of GVHD. CB units have been reported to contain Tregs, but minimal data are available on these cells. Aim of this study was to compare the suppressive functions of Tregs expanded from CB units with those expanded from the peripheral blood (PB) of patients who have undergone an allogeneic SCT. Tregs were purified from mononuclear cells obtained from CB units or PB using the CD4+CD25+ regulatory T-cell isolation kit (Miltenyi Biotec) and expanded for 6 days in 96-well U-Bottom plates coated with anti-CD3 and anti-CD28 MoAbs plus IL-2. To assess the suppressive functions, expanded Tregs from CB units or PB were seeded with naïve autologous effector T cells stimulated with allogeneic dendritic cells (DC) pulsed with apoptotic leukemic blasts, then incubated with [3H]-thymidine and counted in a beta-counter. Suppressor activity was measured as [3H]-thymidine incorporation in the presence or absence of Tregs. The proportion and the immunophenotypic analysis of Tregs present in the CB units (n = 9) -in terms of expression of surface CD4, CD25, CD62L, cytoplasmic CTLA-4 and Foxp3was comparable to those obtained from the PB of allografted patients (n = 9). In addition, Tregs from CB units and from the PB of these patients showed an equivalent expansion capacity [mean fold increase (range), CB units 8.6 (1.8-24); PB patients 9.1 (1.5-16.5) ]. On the contrary, preliminary data show that Tregs expanded from CB units (n = 4) exert a higher suppressive function on the proliferative reaction of T cells stimulated by allogeneic DC compared to Tregs expanded from the PB of allografted patients (n = 2) [mean fold reduction (range), CB units 9.01 (2.66-15.08); PB patients 3.83 (1.49-6.17) ]. Moreover, immunofluorescence analysis demonstrated that Tregs expanded from CB units (n=2) are highly positive for cytoplasmic IL-10 (mean 82%, range 65-99) and negative for IFN-gamma. These results indicate that Tregs contained in CB units exert a potent suppressive function in mixed lymphocyte reaction culture assays and offer further insights into the understanding of the biology of CB transplant.

Double negative tregs are reduced in allo-transplanted patients developing graft-versus-host disease B. Serio* (1), Z. McIver (1), A. Risitano (2) , C. Selleri (2) , J. Maciejeski (1) (1)Cleveland Clinic (Cleveland, US); (2) Federico II University of Naples (Naples, IT) During development of a healthy immune system, central tolerance is induced in the thymus by the negative selection of T lymphocytes that have a high affinity for self antigens. After BMT, central tolerance may be impaired by thymic involution and conditioning regimens resulting in dysregulated alloreactivity. Various subpopulations of regulatory T cells including the Tregs or suppressor T cells with CD4+CD25+FoxP3+ phenotype (FoxP3 Tregs), NK T-cells and the CD3+CD4-CD8-CD56-a/aTCR+ T regs (Double Negative [DN] Tregs) had led to the mechanisms of peripheral tolerance. We investigated behavior of DN following human allogeneic HSCT with regard to the occurrence of graft versus host disease (GvHD) and restoration of T cell receptor (TCR) repertoire. A cohort of 40 patients was investigated; 16 underwent matched unrelated HSCT and 24 received matched sibling grafts. Frequency of DN and TCR repertoire of CD4 and CD8 cells was measured serially and at the time of diagnosis of GvHD by flow cytometry. Our patient population demonstrated skewing of TCR repertoire and we identified a very strong and linear relationship between total number of Vâ family expansions and the grade of GvHD (grade 1-4, N=28, p=.005); this relationship held true when considering separately both the CD4 and CD8 compartments (p=.009 and p=.023, respectively). The median frequency of DN Tregs was calculated to be 0.54% of the total CD3+ T-cell population (range 0.04 -3.11). By using the total number of Vâ family expansions as a gauge of alloreactivity, we observed a reduced number of DN Tregs in those individuals that developed 4 or more Vâ family expansions (N=25, mean 1.16 vs 0.56 cell/ìL, p=.03). In addition we also noted a significant difference in both the mean percentage and mean absolute values of DN Tregs for those individuals that developed GvHD (grade >2) when compared to those that did not (N=29, 0.40 vs 1.32%, p=.004, and 0.8 vs 2.6 cells/mL, p=.024 respectively). The size of DN Treg compartment inversely correlated to the grade of GvHD (grade 1-4) as both percentage and absolute value (N=29, p<.001, and p=.019 respectively). In conclusion we found that Vb family expansion are associated with degree of alloreactvity and GvHD. In addition we show that DN Tregs are reduced after BMT suggesting their important role in peripheral tolerance and alloreactivity. This work contributes to better define the regulatory role of DN and to develop future therapeutic applications.

patients with myelofibrosis. A study of the MDS subcommittee of the Chronic Leukaemia Working Party of the

de Witte on behalf of the MDS Subcommittee of the Chronic Leukemia Working Party of the European Group for Blood and Marrow Transplantation

HLA disparities were 4/6 match (n=49), 5/6 (n=10), and 6/6 (n=1). , respectively. Twenty-one patients received cyclosporine, and 39 patients had tacrolimus alone as GvHD prophylaxis from day C1. Neutrophil engraftment was achieved in 86% (median day; 20). Cumulative incidence of PIR, grade 2-4 and grade 3-4 acute GvHD were 60, 45 and 31%, respectively. Estimated 2-years survival was 47% (95% CI: 34-68%)

Sierra Hospital de la Santa Creu I Sant Pau (Barcelona, ES) Background: Neurological complications (NC) of allogeneic hematopoietic stem cell transplantation (allo-HSCT), involving central (CNS) and/or peripheral nervous system (PNS) are common and remain life-threatening in most cases. Their incidence and characteristics after allo-HSCT have not been well defined

We excluded patients with NC due to CNS relapse of their malignancies as well as encephalopaty in the context of pre-mortem multi-organ failure

CNS complications included seizures in 7 cases, 5 non-focal encephalopaties, 5 meningoencephalitis and 5 strokes or hemorrhages. PNS complications consisted of 4 cases of neuralgia/neuritis and 4 cases of demyelinating axonal neuropathies (1 Guillain-Barré syndrome)

Piñana is supported by grants ISCIII (CM06/00139) O404

Haematopoietic stem cell transplantation for advanced primary MDS in children: results of the EWOG MDS study group B. Strahm* (1)

); (5)Wilhelmina Children´s Hospital

and methotrexate, and additional methylprednisolone in case of a sibling donor. The median clinical follow-up time for the patients was 37 months (3-46 months). Several anti/coagulation activities associated with endothelial cell activation were serially assessed up to 3 months: prothrombin fragments 1+2 (F1+2), thrombin time (TT), FVIII:C, activated protein C-Protein C inhibitor (APC-PCI) complex and protein C act (PC). During conditioning (on day-2) as an early sign of thrombin generation, F1+2 and APC-PCI complex increased 2-3 fold. After engraftment, FVIII:C increased steadily to reach its high maxi-mum on day +24 (273 % ±104 %, median±SD, p<0.001). Interestingly, PC rose (189 % ± 63 %) in parallel with FVIII:C with a 5-fold individual variability. After the engraftment FVIII:C and PC were highly interrelated. GvHD developed in 8 patients and was predicted by early low PC activity (<90%) during conditioning (p=0.007) (OR=16.7). GvHD was also associated with elevated level of F1+2 (> 0.7 nmol/l) (p=0.014) and was predicted by short TT (<15 s) (p=0.004) (OR= 33.1) after the transplantation (on day +10). No patient whose F 1+2 was < 0.7 nmol/l (n=11) developed GvHD. Elevated level of F 1+2 (>1.5 nmol/l) after the transplantation associated with non-relapse mortality. 3 patients with the highest thrombin generation after SCT all died (7-23 months). In con-clusion, early up-regulation of thrombin generation and down-regulation of PC associated with the appearance of GvHD. After allogeneic SCT procedure there is an intimate relation between endothelium regulated coagulation and development of GvHD, suggestive of a new therapeutic target.

Final results from a PHETEMA study L. Rosiñol* (1) , J.J. Lahuerta (2) , A. Sureda (3), J. de la Rubia (4), J. , M. Hernández-García (6), B. Hernández-Ruíz (7), J.A. , J.L. Bello (9), D. Carrera (10), M.J. Peñarrubia (11) , E. Abella (12) , A. León (13), C. Poderós (14) , J.C. J. Besalduch (16) , R. , I. P. Ribas (19), J. San Miguel (8) , J. Bladé (1) (1)Hospital Clinic (Barcelona, ES) ; (2) Background: Two randomized trials showed that tandem transplant result in a significantly longer EFS and OS in patients failing to achieve complete remission (CR) or near-CR with a single transplant. However, other studies failed to show survival benefit from a second transplant but there was no survival plateau. Promising results have been reported using dose-reduced intensity conditioning (Allo-RIC), especially after debulky with an autologous transplant.Aim: To investigate the efficacy in terms of response upgrading and survival from a second transplant intensification in patients with chemosensitive disease who failed to achieve CR or near-CR with a first transplant. Patients and methods: Patients diagnosed with MM from Oct 1999 to Dec 2004 younger than 70 years received 6 courses of VBMCP/VBAD and responding patients were intensified with busulphan/melphalan or MEL-200 followed by stem cell support. Patients not achieving CR or near-CR were planned to undergo a second transplant (second auto with CVBcyclophosphamide, etoposide and BCNU -or MEL-200 intensification or an Allo-RIC with fludarabine/MEL-140 conditioning, if sibling donor available. Results: Eighty-five patients received a second autologous transplant while 26 underwent an allo-RIC. The CR rate was significantly higher with allo-RIC (35% vs. 10%, p= 0.02). There was a trend towards a higher TRM with the allogeneic procedure (5% vs. 15%, p=0.09). After a median follow-up of 51 months for alive patients, there were no significant differences in EFS (48 mos. Vs. nt reached) and OS (80 mos vs. not reached for autologous tandem vs. auto/Allo-RIC). However, there is a plateau in the "Allo-RIC" group beyond 3 years of the second procedure not observed in the autologous arm. A final update will be presented. Conclusions:1) an Allo-RIC transplant after an autologous procedure results in a significantly higher CR rate than a second autologous transplant, and 2) although we found no significant differences in survival between the two transplant modalities, there is a plateau in the allogeneic group. (2) Background: It has been assumed that patients with primary refractory myeloma benefit from early high-dose therapy/stem cell support (HDT. However, in the reported series patients with "unresponsive-progressive disease" vs those "nonresponding-non progressing" were not analyzed separately Aim: Response and survival after early HDT in the two populations of truly primary refractory multiple myeloma (i.e., patients with progressive disease versus those with "no change" or "stable disease" while receiving the initial therapy). Patients and methods: From Oct 1999 to Dec 2004, 829 patients with MM received 6 cycles of VBMCP/VBAD and at least one transplant. 81 of the 829 patients were refractory to VBMCP/VBAD. These resistant patients were scheduled to receive a tandem transplant, the first with Bu-12/MEL-140 or MEL-200 and the second "autologous" with CVB (ciclophosphamide(etoposide/BCNU) or MEL-200 or "allo-RIC" (if donor available) with Fluda/MEL-140. Response and progression were defined by the EBMT criteria. Results: 31 of the 81 primary refractory patients had progressive disease under the initial chemotherapy while 50 Allogeneic stem cell transplantation (SCT) is the only curative treatment approach in patients with myelofibrosis. The major limitation is the high treatment related mortality, which exclude mainly elderly patient from this treatment procedure. To determine the toxicity and efficacy of a dose-reduced conditioning regimen, consisting of busulfan (10 mg/kg), fludarabine (180 mg/m²) and anti-thymocyte globulin (ATG Fresenius: 3x 10 mg/kg for related and 3 x 20 mg/kg for unrelated SCT) followed by allogeneic SCT, we performed a prospective multicenter trial in elderly patients with myelofibrosis in 18 centers in three countries. From 2002 to 2006, 104 patients with a median age of 55 years (r.,32-68) were included and 102 were evaluable for outcome. Risk profile was low risk with constitutional symptoms (18%), intermediate risk (n= 58%) and high risk (n=19%). All but 3 patients received peripheral blood stem cells as graft source either from related (n=31) or unrelated donor (n=69). All but one (1%) patient showed leukocyte and platelet engraftment after a median of 18 and 21 days, respectively. The median duration of leukocyte aplasia was 9 days (r., 3-21). Acute graft-versus host disease (GvHD) grade II o IV occurred in 19% and severe aGvHD III/IV in 7%, while chronic GvHD was seen in 32% of the patients. Non-relapse mortality at 1year was 17% (95% CI: 11-27%) and significantly lower for patients younger than 50 years of age (4% vs 24%, p<0.001) and for patients with low risk vs intermediate/high risk disease (9% vs 26%, p= 0.07), while a higher NRM was seen for patients transplanted from HLA mismatched donors (66 vs 17%, p= 0.006). The cumulative incidence of relapse at 3 year was 25% (95%CI: 15-43%) and influenced by time from diagnosis to SCT of less or more than 24 months (18 vs 40%, p=0.05). Patients with splenectomy had higher incidence of relapse (60 vs 18%, p=0.003). The estimated 3 year overall and event-free survival was 73 and 58%, respectively. The overall survival was influenced by age less than 50 years (92% vs 59%, p=0.009 and low vs intermediate/high risk (94% vs 60%, p=0.03 and HLA mismatch (39 vs 70%, p= 0.02), while no impact on survival was seen for cytogenetic abnormalities, JAK2 mutation status and donor (related vs unrelated) These results of a prospective multicenter study show excellent outcome of a busulafan/fludarabine based reduced conditioning regimen followed by allogeneic stem cell transplantation in patients with myelofibrosis.

Graft rejection after stem cell transplantation following reduced-intensity conditioning is influenced by the underlying disease, the donor type, disease stage and the CD3 content of the graft G.-N. Franke*, A. Mikolajewska, S. Leiblein, H.-K. Al-Ali, E. Hennig, W. Pönisch, D. Niederwieser, T. Lange University of Leipzig (Leipzig, DE) Objectives: Stem cell transplantation (SCT) after reduced intensity conditioning (RIC) is routinely used as a curative approach for older and medical impaired patients with haematological malignancies. In contrast to SCT with conventional conditioning, graft rejection (GR) remains an important issue. We analyzed patients with SCT after 2 Gy total body irradiation (TBI) with or without fludarabine (FLU) conditioning to identify risk factors for GR. Patients and methods: 330 patients with a median age of 58 (range 17 -74) years, underwent allogeneic SCT (BM=11, PBSC=319) from a related (n=115) or unrelated (n=215) donor for AML (n=121), CML (n=34), NHL/MM (n=102), MDS/MPS (n=38) or other diseases (n=35). Conditioning regimen consisted of 2 Gy TBI at day 0 and FLU 30 mg/m2 from -4 to day -2 (n=305) followed by Cyclosporin A and Mycophenolate Mofetil. Results: 28 (8.5%) patients developed primary (n=23) or secondary graft failure (n=5) defined as a T-cell chimerism of <10% donor cells. In univariate analysis and also in multivariate analysis, diagnosis of CML (p=0.011), unrelated donor (p=0.029), early disease stage prior to SCT (p=0.048) and low CD3 cells in the graft (p=0.002) were identified as independent predictors for GR. The relative risk (RR) to experience GR was 4.8, 4.2, 3.3 and 1.8, respectively. Even in a subgroup analysis with PBSC recipients only, CML, unrelated donor and disease stage were associated with higher risk of GR. Furthermore, a lower CD3 count increases the risk of GR by a RR 1.5 (p=0.06) in the multivariate model (median 3.2x108/kg BW). An increase in CD3 cells was not associated with increased incidence of acute GvHD grad II-IV until day 100 (p=0.935). In contrast, the CD34-content of the graft had no impact on GR either as categorical (>median vs. <median) or as continuous variable (p=0.454). Conclusion: In contrast to given risk factors for graft rejection, the T-cell content of the graft can be individually requested in every patient. Since higher CD3 cells in the graft reduces the risk of GR without subsequent acute GVHD in our cohort of patients, we recommend the administration of CD3 cells >3.2x108/kg BW especially in patients with CML, unrelated donor and early disease stage in RIC-SCT.

Unrelated cord blood transplantation for adult patients with acute myeloid leukaemia/myelodysplastic syndrome using a reduced-intensity conditioning regimen consisting of fludarabine, melphalan and total body irradiation K. Masuoka*, K. Ishiwata, M. Tsuji, S. Takagi, H. Yamamoto, D. Katoh, Y. Matsuhashi, S. Seo, N. Matsuno, N. Uchida, A. Wake, S. Miyakoshi, S. Taniguchi Toranomon Hospital (Tokyo, JP) Objectives: To evaluate the efficacy of unrelated cord blood transplantation (UCBT) in a reduced intensity (RI) conditioning regimen consisting of fludarabine (125mg/m²), melphalan (80mg/m²) and total body irradiation (400 cGy), we analyzed retrospectively the results of 60 adult patients with acute myeloid leukemia (AML)/ myelodysplastic syndrome (MDS) in our hospital. Patients and Methods: We reviewed medical records of 60 patients with AML/MDS who had received single cord blood unit between November 2003 and July 2007 at Toranomon Acute leukaemia 2 / Myelodysplasia O397 Myeloablative allogeneic stem cell transplantation with unrelated donors for high-risk acute myeloid leukaemia: no increase in relapse with alemtuzumab-depleted grafts P. Kottaridis* (1), K. Thomson (2) 

The role of T cell depletion in myeloablative allogeneic stem cell transplantation for acute myeloid leukemia (AML) remains uncertain. Dose intensification may compensate for any potential loss of graft-versus-leukemia effect, and the use of T cell replete grafts is associated with significant morbidity and mortality, particularly when using unrelated donors. This study describes the results in 44 patients with high-risk AML transplanted with an alemtuzumab-containing myeloablative regimen, using unrelated donors. Median age at transplant was 34 years and 20 patients (45%) had donors mismatched at 1-3 HLA loci. All patients were high-risk for relapse, as defined by induction failure, adverse cytogenetics, >CR1, or secondary disease (4 preceding myelodysplastic syndrome [MDS] , 3 therapy-related). Two patients had untreated relapse at the time of transplant, and 9 of the remaining 42 (21%) had refractory disease. The conditioning regimen was cyclophosphamide 60mg/kg for 2 days, fludarabine 30mg/m² for 3 days and Total Body Irradiation (TBI) 14.4Gy in 8 fractions over 4 days. Stem cell source was peripheral blood in 37 and bone marrow in 7, and graft-versus-host disease (GVHD) prophylaxis was with 20mg alemtuzumab added to the stem cells prior to infusion and cyclosporin 3mg/kg. Median follow-up for surviving patients was 36 months. Estimated event-free survival (EFS) was 60% at 1yr and 55% at 4 years. Acute GvHD grade II occurred in 11% (no grade III-IV) and extensive chronic GvHD in 23%. Non-relapse mortality (NRM) was 19% at 4 years with no events beyond 10 months post-transplant, and estimated relapse risk was 27% at 4 years with no events beyond 15 months. For EFS and relapse risk, the only significant variable was chemosensitivity pre-transplant, with inferior EFS (65% at 4 years for chemosensitive patients vs 27% for chemorefractory, p=0.026) and worse relapse risk (15% at 4 years for chemosensitive patients vs 55% for chemorefractory, p=0.010) in those not in complete remission at transplant. There was no impact on NRM or relapse risk depending on the presence of HLA mismatch, or of significant GVHD (>grade I acute, or any chronic GVHD). Use of this regimen therefore permits the successful transplantation of patients with high-risk disease and HLA-matched/mismatched unrelated donors, with minimal acute GvHD, relatively low NRM and no evidence of an excessive relapse rate, particularly in those in complete remission at transplant.

Haematopoietic Stem Cell Transplants (HSCT) using Unrelated Donors (UD) has become an important and viable option in the treatment of Acute Leukaemia (AL). We have previously shown an increased risk of relapse with HLA-DPB1 matching and independently, with NOD2/CARD15 genotype. In light of these data, we have analysed a larger UD-HSCT cohort in order to establish the impact on outcome when both variables are considered. HLA and NOD2/CARD15 typing was performed on 304 AL UD-HSCT pairs. Transplants were between 1996 and 2005. Diagnoses were ALL (47%) and AML (53%). 67% of the cohort were 10/10 HLA matched and 16% were also HLA-DPB1 matched. Myeloablative conditioning regimens were used in 74% of transplants. 82% of conditioning protocols included T-cell depletion. Bone marrow was used in 72% of transplants, the remaining 28% using peripheral blood stem cells. Two forms of post-transplant immunosuppression predominated, Cyclosporine A and Methotrexate (47%) and Cyclosporine A alone (38%). Based on our previous data, the cohort was grouped according to their relapse risk, group 1 (DPB1 matched; NOD2/CARD15 SNP, n=24), group 2 (DPB1 matched; NOD2/CARD15 Wild-Type (WT) or DPB1 mismatched; NOD2/CARD15 SNP, n=112) and group 3 (DPB1 mismatched; NOD2/CARD15 WT, n=168). Disease relapse differed significantly between the three groups (1 year: group 1 68%, group 2 48%, group 3 30%, p=0.006). This finding persisted in multivariate analysis where being in either group 2 or 3 was protective towards relapse as compared to group 1 (RR 0.321; p=0.001 and RR 0.478; p=0.031 respectively). In group 1 (high risk), this resulted in decreased Overall Survival (OS) (33% vs 54% in group 3, RR 0.617; p=0.080). The best OS was seen in group 3 (low risk) where in addition to low relapse, there was increased acute and chronic Graft-versus-Host Disease (GvHD) (p=0.0019 and p=0.002 respectively). In this cohort, limited cGvHD was S74 associated with reduced relapse (p=0.01) and better OS (p<0.0001). In accordance with our theory, being HLA-DPB1 matched and NOD2/CARD15 SNP predicts for the worst outcome with significantly increased relapse and reduced OS. The ideal pairing is HLA-DPB1 mismatched and NOD2/CARD15 WT. These data suggest that prospectively typing AL patients for both variables will allow the prediction of transplant outcome and will allow the effects of being independently HLA-DPB1 matched or NOD2/CARD15 SNP to be offset by intelligently selecting a suitable, less precarious donor. 8 (t8) is the most common chromosomal abnormality in AML. The prognostic impact of t8 as a sole aberration in AML remains unclear, conferring either an intermediate or a poor prognosis. Indeed, patients with t8 occurring with other cytogenetics abnormalities seem to have the prognosis conferred by the accompanying aberration. The aim of this study was to describe the results of allogeneic transplantation in a large series of AML patients exhibiting isolated or associated t8 and to compare outcome with intermediate risk AML patients exhibiting normal caryotype and receiving allo-HSCT. 182 AML patients were identified (males n=100; females n=82) with isolated (n=136) or associated (n=46, favourable group n=8, intermediate n=30; high-risk n=7; unknown n=1) t8, allografted with an HLA identical sibling (n=113) or an unrelated donor (n=69) between 1990 and 2007 and reported to the EBMT. Median age was 37 years (range: 17-68). Median interval between allo-HSCT and diagnosis was 165 days (range: 71-946). The proportion of patients in CR1, CR2/CR3 or active disease was 59%, 13% and 28%. Myeloablative and reduced intensity conditionings were performed in 148 and 34 patients respectively. GVHD prophylaxis consisted of CsA/Methotrexate or CsA/MMF in 49%. Engraftment was observed in 171 patients (95%). Grade II and III/IV acute GVHD occurred in 21% and 11%. Chronic GVHD developed in 45%. With a median follow-up of 48 months (range: 3-180), 5-year LFS, relapse rate and NRM were 45%, 30% and 25%. Status at transplant (CR vs others) (p<0.0001, HR 0.3, 95% CI: 0.2-0.46), female sexe (p=0.03, HR 1.61, 95% CI: 1.05-2.48) and HLA sibling donor (p=0.02, HR 1.64, 95% CI: 1.08-2.49) were significant predictors for better LFS. 5-year LFS was similar between AML patients with isolated or associated t8 (41% vs 55%, p=0.11). Whenconsidering only patients allografted in CR1, 5-year LFS was similar between isolated t8 AML patients (55%, n= 82, median follow-up 51 months), associated t8 AML patients (55%, n=26, median follow-up 51 months) and AML patients with normal caryotype (58%, n=1782, median follow-up 36 months). We conclude that allogeneic transplantation in first linetherapy is a valid therapeutic option in patients exhibiting isolated or associated t8. Isolated or associated t8 do not confer bad prognosis and AML patients exhibiting such aberrations have to be considered as intermediate risk patients as they likely have the same outcome of AML patients with normal karyotype allografted in CR1.

Regimen intensity in acute myelogenous leukaemia: addition of 400cGy total body irradiation to a myeloablative fludarabine/busulphan/thymoglobulin allogeneic transplant regimen reduces relapse without increasing transplant-related mortality J. Russell* (1), W. Irish (2) , L. Savoie (1) Some attempts to intensify myeloablative stem cell transplant (SCT) conditioning protocols for acute myelogenous leukaemia (AML) have reduced relapse only at the expense of increased transplant-related mortality (TRM). A regimen of intravenous busulphan, fludarabine and Thymoglobulin has been well tolerated but followed by a substantial relapse rate. We report the results of a study to enhance the antileukaemic effect of this protocol by adding a low dose of total body irradiation (TBI). 179 patients were treated between 1999 and 2006 with fludarabine 50mg/m 2 daily x 5 and intravenous busulphan 3.2 mg/kg daily x 4. 88 had additional total body irradiation (TBI) 200cGy x 2 on day -1 or 0. Graft-versus-host disease (GvHD) prophylaxis was cyclosporine A, methotrexate and Thymoglobulin (Genzyme) 4.5 mg/kg total dose. Median age was 46 (range 18-66) years in TBI recipients, 42 (16-65) for the non-TBI group. There was no difference in the proportions of SCT with good risk (GR, CR1 or CR2) recipients (65% vs 56%), alternative (unrelated or mismatched related) donors (49% vs 40%) CMV +ve donor +/or recipient (60% vs 68%), female donors to male recipients (18% vs 19%) and high risk (HR) recipient cytogenetics (20% vs 14%) between the TBI and no TBI groups respectively. More TBI recipients received blood cells (91% vs 74%, p = 0.001) and consequently higher CD34+ cell doses (median 5.9x10 6 /kg, range 0.75-17.69 vs median 4.310 6 /kg, range 0.64-23.87, p <0.0001). Follow-up of survivors was 12-83 months (median 31) for TBI recipients and 13-100 months (median 79) for those not given TBI. There was no difference in incidence of acute GvHD grade II-IV at 23% vs 16±4%, acute GvHD grade III-IV at 9% vs 9% and chronic GvHD at 55% vs 64% with and without TBI respectively. Outcomes at 3 years are shown in the table: After adjusting for all the above risk factors the Cox proportional hazard ratio for relapse was 0.32 (95% CI 0.17-0.6) in favour of TBI (p = 0.004). There was no effect of GvHD on relapse. The impact of TBI on relapse without affecting TRM resulted in a significantly decreased risk of mortality with a hazard ratio of 0.56 (0.33-0.93, p = 0.03). The hazard ratio for disease-free survival was 0.46 (0.28-0.77, p = 0.003). This regimen allows some intensification by adding a low dose of TBI without an effect on TRM but a reduction in relapse, confirming the importance of regimen intensity in transplantation for AML. S75 O401 Long-term survival in patients suffering from AML with a complex aberrant karyotype after early allogeneic stem cell transplantation using the FLAMSA-RIC regimen: results from a prospective phase II trial C. Schmid* (1), M. Schleuning (2) Introduction: Patients suffering from acute myeloid leukemia (AML) with a complex aberrant karyotype (i.e. ≥ 3 cytogenetic aberrations) usually show poor response to chemotherapy and have a grim prognosis. Allogeneic stem cell transplanttaion (alloSCT) is the treatment of choice, although the unfavorable prognostic value of a complex karyotype was preserved in most studies. In contrast, in the FLAMSA-RIC pilot trial, a post-hoc subgroup analysis revealed similar results for patients with a complex karyotype as compared to more favorable cytogenetic subgroups (Schmid, Schleuning et al, JCO 2005) . Therefore, a prospective phase II trial for patients with complex karyotype AML was initiated within the German AML cooperative group, to evaluate the role of alloSCT, performed as early as possible after diagnosis. The FLAMSA-RIC preparative regimen contained the sequence of intensive cytoreductive chemotherapy (FLAMSA), followed three days later by reduced intensity conditioning (RIC; 4Gy TBI; cyclophosphamide, ATG). Patients and Methods: 20 patients from 4 centers (median age: 50,1, range 20-63 years), with a median of 7 (range: 3-15) cytogenetic aberrations were included. Median time from diagnosis to transplantation was 91 days. Eight patiens had received one, 12 had received two courses of conventional chemotherapy. Stage at start of FLAMSA-RIC was CR1 in 7, first cytogenetic relapse in 1, and primary induction failure in 12 patients, including 8 with persistent disease after high-dose AraC. Donors were HLA-identical siblings in 9, matched unrelated donors in 8, and 1-AG-mismatched unrelated donors in 3 cases. Results: Nineteen patients engraftet (median day: +18), one died in aplasia. 17 patients achieved molecular CR, 1 regenerated with blasts, and one had cytogenetically persistent disease. AGvHD developed in 13 patients, but reached > grade II only in 3. Five patients developed cGvHD. Relapse occurred in 2 patients. After a median follow up among survivors of 23 (range: 3-33) months, 8 patients have died from leukemia (n=3) or treatment-related causes (n=5). Overall survival at one and two years from transplantation is 74% and 58%, the corresponding leukemia free survival is 69% and 58%. Conclusion: To our knowledge, this is the first trial specifically addressing patients with a complex aberrant karyotype. Although the numbers are small, the results suggest a promising activity of early alloSCT in this otherwise very unfavorable subgroup of patients with AML. The iron (Fe) chelator, desferrioxamine (DFO), has been shown to have both antiproliferative and apoptotic effects in tumor cells. Fe depletion results in G1/S cycle arrest and cell apoptosis and DFO acts as hypoxia-mimetic agent by accumulating the hypoxia-inducible factor-1 alpha protein which regulates the cellular response to hypoxia by cessation of growth. In this retrospective study we evaluated the effect of DFO administration and iron overload (IRO) in relapse incidence in 127 consecutive patients (pts), allografted for malignant diseases (myeloid: 76, lymphoid: 51). The disease phase was early in 45, intermediate in 38 and advanced in 44 pts. Thirty received non ablative and 97 ablative conditioning. Peripheral blood (116) or bone marrow (11) grafts were donated from 101 siblings, 23 matched unrelated and 3 haploidentical donors. Graft vs Host disease (GvHD) prophylaxis was consisted of Cyclosporine or Prograf plus Methotrexate. Non-responders or relapsed pts within 2 months post allotransplant, were excluded. According to our center policy, pts with established engraftment and IRO, as evidenced by ferritin levels>2000ng/ml, elevated liver enzymes or/and liver MRI indicating hemosiderosis, receive DFO. Among 95/127pts with IRO, 31 were treated with DFO (50mg/kg x 5days/week, iv or sc for 2 months at least). The 5year relapse rate (RR) was significantly lower in pts treated with DFO than in non-treated pts (22% vs. 53%, p=0,003) and this benefit was restricted to myeloid malignancies only. In a multivariate analysis we examined DFO, chronic and acute GvHD, disease phase, graft source, type of donor, origin of malignancy and preparative regimen as risk factors for relapse. DFO administration retained its significance (p=0,02) while the absence of cGvHD and the advanced disease phase were significant factors for relapse (p<0,03). In order to explore whether IRO affects the relapse incidence we separately examined the Fe-overloaded pts who received no chelation (64) vs the non-Fe-overloaded pts (32). The 5-year RR for untreated pts was 60% vs 33% for those with ferritin values <2000ng/ml (p=0,04). IRO remained as significant risk factor for relapse in the multivariate analysis (p=0,03). This is the first clinical study to investigate the role of DFO in relapse incidence post allotransplant suggesting a possible role for DFO therapy post-transplant. Prospective studies are needed to clarify if DFO may have a role in relapse prevention.

Equivalent disease-free survival results after CBT and BMT from unrelated donor using TBI containing myeloablative regimen and calsinulin inhibitors plus MTX methods in patients with MDS in Japan: multivariate analysis by competing risk regression models S. Takahashi*, T. Yamaguchi, M. Monna-Ooiwa, S. Taniguchi, H. Akiyama, T. Morii, Y. Nagari, Y. Takaue, S. Okamoto, K. Miyamura, H. Sao, T. Nagamura, S. Kato, T. Kawase, Y. 

The result of single institutional analysis in Japan has shown cord blood transplantation (CBT) is promising in adults with hematologic malignancies including myelodysplastic syndrome (MDS) and is comparable with bone marrow transplantation (BMT) from unrelated or related donors. We evaluated safety and efficacy of both BMT and CBT from unrelated donors using the data of MDS patients within the Japan Marrow Donor Program and the Japan Cord Blood Bank Network registry database and related those to biological and procedural factors. Clinical data of 965 patients with MDS including transformed acute myelogenous leukemia who received unrelated BMT (n=532) or unrelated CBT (n=433) between 1993 and 2006 were collected. The median periods of follow-up for survivors were 21 months for BMT and 12 months for CBT. We analyzed the hematopoietic recovery, incidences of graft-versus-host disease (GVHD), risks of transplant-related mortality (TRM), relapse and disease-free survival (DFS) using competing risk regression models.Advanced primary myelodysplastic syndrome (MDS) represents a subgroup of childhood MDS characterized by bone marrow (BM) dysplasia and an increased blast count in peripheral blood (PB) and/or BM. Allogeneic hematopoietic stem cell transplantation (HSCT) is the only curative treatment. Here we report the results of 105 patients (pts) (71 males/34 females) enrolled in the prospective EWOG MDS study. According to the highest WHO type prior to HSCT pts were classified as RAEB (59), RAEB-T (30) and MDR-AML (16). Median age at diagnosis was 10.6 yrs (1.0-18.2) and the median time from diagnosis of advanced MDS to HSCT was 4.1 mo (1.0-31.2) . Cytogenetics revealed monosomy 7 in 37 pts, a complex karyotype in 11 pts, other abnormalities in 22 pts, no abnormalities in 31 pts and was unknown in 4 pts. Intensive chemotherapy was given to 31 pts prior to HSCT. 43 children were transplanted from an HLA matched family donor (MFD), while the remaining 62 were given an allograft from a matched or 1 Ag mismatched unrelated donor (UD). The preparative regimen included busulfan, cyclophosphamid and melphalan in all cases. Stem cell source was unmanipulated BM, PB and cord blood in 73, 30 and 2 pts, respectively. Cyclosporine-A alone was used as graft versus host disease (GVHD) prophylaxis in most children transplanted from a MFD, while the majority of pts transplanted from an UD were given Cs-A, MTX and anti-thymocyte globulin. With a median follow up of 3.1 yrs (0.1-10.2) 62 out of 105 pts are alive and disease free, 19 experienced relapse, and 24 died of transplant related complications. The 5-year probability of event-free survival (EFS) is 0.56 [0.46-0.66] while the 5-year cumulative incidence of transplant-related mortality (TRM) and disease recurrence are 0.23 [0.16-0.33] and 0.21 [0.14-0.32], respectively. The cumulative incidence of grade II-IV GVHD is 0.47 [0.37-0.57]. There is no significant difference in EFS, TRM and disease recurrence according to donor, MDS subtype, karyotype or the treatment applied prior to HSCT. Age at HSCT >12 yrs, an interval from diagnosis of advanced MDS to HSCT ≥ 4 mo and the presence of GVHD contribute to a significant increase in TRM. These results indicate that a large proportion of children with advanced MDS can be rescued by allogeneic HSCT. TRM remains a main cause of treatment failure especially for children older than 12 yrs. For these patients new strategies to reduce TRM will have to be defined.

Introduction: Patients with a history of an allogeneic haematopoietic stem cell transplant(HSCT)have an increasing risk to develop a secondary cancer.The association of the HLA system to cancer is well known.Sibling donors share the same HLA alleles, and the question is,if they have an increased tumour incidence compared to the general population.This is of concern because the possibility of a tumour induction as a result of mobilisation with G-CSF has been discussed. Methods: All HLA-identical sibling pairs with an HSCT from 1974 to 2001 for a malignant disease and living in Switzerland were evaluated. General data of the donors were taken from the patient's charts of the transplant unit.Donors were contacted per call or mail and asked about their current health status.In case of a malignant cancer,information on date of diagnosis,localisation,treatment was obtained.Data were compared with the age-,and sex adapted cancer incidence rate of the Swiss association of cancer registries(SACR).The same information was retrieved for the patients. Results: 318 pairs were identified,in 291(92%)the donors (142 men(m),149 women(w))could be contacted. Median observation time was 13.8 years(y)(range 5-32y).146 donors were <50y,89 between 50-59y,29 between 60-64y,20 between 65-69 y and 7 >70y old, hence 85% were <60y old.Seventeen (6%)donors,12 bone marrow and 5 peripheral blood donors,had developed a total of 18 cancer of 9 different localisations (mamma,prostate,skin,bone marrow, colon,bronchus,stomach,bladder,ORL).According to the incidence rate of the SACR,3.3 tumours in m and 6.8 in w would have been expected,3 (m) (RR0.91) and 4 tumours (w) (RR 0.84) were found in donors between 0-49y.In the age category 50-69 y,4.5 tumours in m and 4.8 in w were expected and 5(RR1.11)respective 6(RR 1.23)observed.In the subgroup of men between 60-64y 1.07 tumour were expected and 4 S77 observed (p<0.02).No donor >70y developed a tumour.4 of the 291 donors had died,3 from the tumour,1 of cardiac disease.In 12 patients a secondary tumour was diagnosed post HSCT. Conclusion: We observed a definite number of donors with a malignancy after HSC donation. Absolute numbers were similar to the numbers of tumours in their transplanted siblings, observed versus expected rates were similar to an age and sex matched population except in male donors between 60-64 y by now. Data from this single centre cohort remain yet inconclusive but underline the need for international collaborative donor follow up. The number of transplants is consistently increasing for AML, ALL, and lymphoid malignancy in all countries/regions except for Vietnam. Meanwhile, the number of HSCT for nonmalignant diseases such as aplastic anemia and hemoglobinopathy has been stable in all countries except for Iran, in which country it is recently increasing. The proportion of hemoglobinopathy in all HSCT differed widely among countries; 0% in Japan and China, 2% in Singapore, 3% in Taiwan, 5% in Hong Kong, 10% in Malaysia, 13% in Vietnam, 22% in Iran, (all HSCT until 2006, only 2006 in Taiwan) . The trends changed quite differently among countries and regions over time for other diseases, most strikingly in CML. In Japan, the number of HSCT for CML had consistently increased until 2000 (n=307 in 2000) and then rapidly decreased due to introduction of imatinib (n=98 in 2005). In contrast, it is dramatically increasing in China (n=8 in 2000 China (n=8 in to n=98 in 2005 and stable in other countries. The number of HSCT for MDS and multiple myeloma is consistently increasing only in Japan (and Iran for myeloma), but not in other countries. HSCT for solid tumors had been commonly performed only in Japan, but the number has decreased since 1997 (n=337 in 1997, n=163 in 2005) . In summary, HSCT is developing in most of the Asian countries/regions. However, disease indications and trends differed widely among them probably due to different economics situations, health service systems, and availability of donors and agents such as imatinib for CML.

Short-and long-term side effects in the healthy donors of allogeneic haemopoietic peripheral cells mobilised with lenograstim: a single-centre experience M. Martino*, G. Console, E. Massara, E. Spiniello, I. Callea, F. Gatto, G. Messina, T. Moscato, R. Fedele, G. Irrera, P. Iacopino U.O. Ematologia con Trapianto (Reggio Calabria, IT) Healthy allogeneic donors, who were mobilized with lenograstim and underwent hemopoietic peripheral cells (HPC-A) collection at our Institution, were enrolled in a shortand long-term surveillance protocol for a 10 year period. To date, 171 donors have been assessed with a median followup of 55 months (2 -145) ; for 71 subjects, the follow-up is > 60 months and for 20 subjects is > 96 months. Healthy donors received lenograstim at a median dose of 9.9 µg/kg (range 7-15). Bone pain was reported as the most common adverse event (70.2 % of donors). Common associated symptoms included fatigue (16.4 %), fevers (5 % ), headache (26.3 %), nausea (12.9%), insomnia (16.4 %). Spleen size increased in 4.1 % of donors (> 2 cm exceeding the marginal cost at physical examination ). All donors experienced side effect resolution within 2-4 days of lenograstim discontinuation. Leukocyte mean peak values were 48 x 109/L and the nadir of platelet counts reached mean values of 93 x 109/L and; however, such a decrease was not complicated by bleeding manifestations. The hemoglobin concentration decreased slightly but not significantly. Leukocyte and platelet counts returned to normal values in about one week. No vascular disorders and cardiac disease occurred. Long-term observation included adverse events in donors after 30 days from HPC-A mobilization and any neoplastic or not disease developed any time post donation. 4 donors showed persistent, slight leucocytopenia until the second month, with recovery in the fourth month of follow-up. 18 donors showed an AST and ALT result 2.5 times the upper limit of normal until the second months of follow-up. Transit ischemic attack occurred in 1 donor, (39 months post after donation). 1 autoimmune event has been reported 28 months post-G-CSF (anckylosing spondylitis); 1 donor with a history of chronic obstructive pulmonary disease developed a secondary polyglobulia (50 months post-G-CSF); 1 donor developed a gastric tumor, 19 months post-donation. No hematological malignancy was observed. In conclusion, our main findings are that the primary toxicity of G-CSF administration is bone pain and that no cardiovascular events was related to the donation. With a median follow-up near to 5 years, no hematological disease was observed in our cohort of donors.

Effect of the search of an unrelated stem cell donor in patients with high-risk leukaemia: prospective, singlecentre study on intention to treat analysis A.P. Iori* (1), V. Valle (1) Allogeneic stem cell transplant (HSCT) plays a major role in the treatment of acute leukemia (AL) patients with high risk (HR) features at diagnosis or in ≥ 2nd complete remission (CR). Between 1995 and 2007, 164 patients -median age 12 years (1 -59) -with HR AL followed at our Center and lacking a family HLA compatible donor were addressed to a search of HSC donor through the BMWW Registry and Cord Blood Banks. The aim of this prospective study was to assess the effect of the search on the outcome of patients with HR AL on an intention to treat analysis. Forty-three patients started the S78 search in I CR, 84 in II CR, 4 in >II CR, 33 in relapse. Fifty-five % of the 131 patients who entered into the study in CR showed a disease relapse at a median of 4 (1-20) and 2 (1) (2) (3) (4) (5) (6) (7) (8) months from the start of the search for patients in I-II CR and > II CR, respectively. Sixty-nine patients (42%) underwent an HSCT, 32% in a more advanced phase compared to the start of the search. Nineteen of the 33 patients (57%) who started the search in relapse obtained a CR and underwent an HSCT. Globally, 45% of patients failed to undergo a transplant because of lost eligibility due to disease progression. For the entire population, the 12 year survival probability and DFS were 19%. Disease progression was the major cause of death. When the variables affecting outcome parameters were analyzed in univariate analysis, the occurrence of relapse during the search period and the transplant procedure affected OS (p=0.001) and DFS (p<0.0001). A more advanced phase of the disease at the start of the search and no transplantation were the factors negatively affected the relapse (p=0.01, p< 0.0001). For patients who underwent an HSCT, the factors which negatively affected OS and DFS were a relapse during the search and a more advanced disease phase at transplant. Both these factors plus the disease phase at the start of the search affected the relapse for transplanted patients. In conclusion, by decreasing the length of the search (4 months for patients in I-II CR and 2 months for patients with a more advanced phase of the disease) the risk of relapse can be reduced, thus increasing the possibility of carrying out a transplant and the transplant success. Moreover, starting the search in relapse we may obtain the CR while waiting for the HSCT. Therefore, during the search of an HSC donor the "timing" of the transplant must be considered the major strategic factor for patients with HR AL. The goal of this prospective study was to evaluate the impact of donor's and recipient's genetic polymorphisms regarding components of innate immunity on outcome of alloHSCT. 102 consecutive patients with hematological malignancies, aged 32(18-58)y, treated with alloHSCT from either sibling (n=34) or matched unrelated (n=68) donors were included. The conditioning was myeloablative. GVHD prophylaxis consisted of CsA, Mtx ± ATG. Donors and recipients were tested for single nucleotide polymorphisms(SNP)8,12,13 of the NOD2/CARD15 gene, TLR4(299), TLR4(399), TLR5(stop codon C1174T) and IL23R(11209026), as well as KIR genotype. In addition, immune reconstitution was studied. OS rate at 2y was significantly lower in alloHSCT with at least one activating KIR mismatch compared to transplants with full compatibility (62%vs.86%, p=.01). In particular, the presence activating killer immunoglobulin-like receptors (KIR) in the donor with its absence in the recipient (D+R-) was associated with decreased rates of OS (60%vs.78%, p=.01) and DFS (58%vs.82%, p=.005), as well as increased NRM (27%vs.7%, p=.01). KIR2DS1 and KIR3DS1 D+R-mismatches resulted in increased risk of grade II-IV acute GvHD, while KIR2DS3 and KIR2DS2 D+R-mismatches were associated with increased risk of chronic GvHD. D+R-activating KIR mismatches correlated with increased CD8+/CD4+ T cell ratio up to day +100. In all cases of incompatibility regarding KIR2DS1, KIR2DS2 and KIR3DS1, T cells with expression of respective receptors could be detected up to 360 days after alloHSCT. The presence of SNP8 of the NOD2/CARD15 gene in the recipient was associated with decreased probability of OS (20%vs.70%, p=.005) and DFS (20%vs.70%, p=.01) as well as increased NRM (60%vs.17%, p=.01) and grade III-IV acute GvHD (67%vs.8%, p=0.02). In a multivariate analysis adjusted for other potential risk factors, increasing number of D+R-activating KIR mismatches as a continues variable appeared to independently influence OS, DFS, NRM, grade II-IV acute GvHD, and chronic GvHD. Recipient SNP8 of NOD2/CARD15 was predictive for OS, DFS, NRM, grade III-IV acute GvHD, and chronic GvHD. SNPs of TLR and IL23R genes had no impact on outcome. Conclusions: Both activating KIR D+R-mismatches and recipient SNP8 of NOD2/CARD15 appear to enhance alloreactivity and independently influence survival after alloHSCT. Evaluation of these polymorphisms may contribute to better donor selection and optimization of the alloHCT procedure. The interpretation of the role of HLA-DPB1 in unrelated haematopoietic stem cell transplantation is subject to discussion. We have investigated the role of HLA-DPB1 allele matching on haematopoietic stem cell transplantation (HSCT) outcomes in 161 recipients who were HLA -A, B, C, DRB1, DQB1 matched with their unrelated donors at the allelic level (10/10). Additionally, we analysed the association of polymorphic amino acids mismatches of DPB1 molecule with HSCT endpoints, and the permissiveness concept of Zino and colleagues (Zino et al., 2004) . DPB1 allele mismatches were significantly associated with an increased incidence of acute graft-versus-host disease (aGvHD), transplant-related mortality (TRM) and worse overall survival (OS). We observed that the mismatch at amino acid position 69 increased the risk for both aGvHD and TRM. Risk factors for aGvHD also included mismatches at positions 8, 9, 35, 76 and 84 . This is, to our knowledge, the first report of an in vivo effect of single amino acid mismatches on HSCT outcomes. In our study, grouping of allelic mismatches into permissive and nonpermissive categories and their association with transplantation endpoints proved relevant for TRM but not for other clinical endpoints. G.F. Torelli*, R. Maggio, N. Peragine, M.S. De Propris, B. Lucarelli, M.G. Mascolo, M. Screnci, S. Salvatori, L. Malandruccolo, A.P. Iori, A. Guarini, R. Foà Sapienza University (Rome, IT) Studies in mouse models of stem cell transplant (SCT) have shown that the infusion of culture-expanded regulatory T cells (Tregs) can be effective in preventing and suppressing GVHD, while still allowing a GVL effect. Cord blood (CB) stem cells are now broadly used in the unrelated SCT setting and